From b5703024@cc.ntu.edu.tw Sat Jan 1 05:51:17 2000 From: b5703024@cc.ntu.edu.tw (蕂) Date: Sat, 1 Jan 2000 13:51:17 +0800 Subject: check the destination address in ns Message-ID: <002601b89a6a$96f5f140$8c14708c@r7942050> hello everyone: Recently , my teacher want me to add some code in ns(in C++ field) could I check the destination address of a packet in link module??how to do it?? thanks wbliu From lhyang@cti.com.cn Mon Jan 3 03:41:00 2000 From: lhyang@cti.com.cn (yanglihai) Date: Mon, 3 Jan 2000 11:41:00 +0800 Subject: question of xgraph using? Message-ID: <001701bf559c$5a6cc480$350279c8@jimyang> This is a multi-part message in MIME format. ------=_NextPart_000_0014_01BF55DF.681C0D00 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 aGkgZXZlcnlvbmUsDQogICAgd2hvIGNhbiB0ZWxsIG1lIGhvdyB1c2UgeGdyYXBoIHRvIHNob3cg cXVldWUgYW5kIGRyb3AgaW5mb3JtYXRpb24/DQppdCBvbmx5IGhhcyBmb3JtYXQgb2YgdHdvIGRp bWVudGlvbnMgYmVpbmcgcmVhZCBieSB4Z3JhcGgsDQpidXQgSSBkb24ndCBrbm93IGhvdyB0byBn ZW5lcmF0ZSB0aGUgZm9ybWF0IGRhdGEsDQphbnkgaGVscCBpcyBhcHByZWNpYXRlZC4NCg0KICAg ICAgICAgICAgICAgICAgICBKaW0gWWFuZw0K ------=_NextPart_000_0014_01BF55DF.681C0D00 Content-Type: text/html; charset="gb2312" Content-Transfer-Encoding: base64 PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWdi MjMxMiIgaHR0cC1lcXVpdj1Db250ZW50LVR5cGU+DQo8TUVUQSBjb250ZW50PSJNU0hUTUwgNS4w MC4yMDE0LjIxMCIgbmFtZT1HRU5FUkFUT1I+DQo8U1RZTEU+PC9TVFlMRT4NCjwvSEVBRD4NCjxC T0RZIGJnQ29sb3I9I2ZmZmZmZj4NCjxESVY+PEZPTlQgZmFjZT0iVGltZXMgTmV3IFJvbWFuIj5o aSBldmVyeW9uZSw8L0ZPTlQ+PC9ESVY+DQo8RElWPjxGT05UIGZhY2U9IlRpbWVzIE5ldyBSb21h biI+Jm5ic3A7Jm5ic3A7Jm5ic3A7IHdobyBjYW4gdGVsbCBtZSBob3cgdXNlIA0KeGdyYXBoIHRv IHNob3cgcXVldWUgYW5kIGRyb3AgaW5mb3JtYXRpb24/PC9GT05UPjwvRElWPg0KPERJVj48Rk9O VCBmYWNlPSJUaW1lcyBOZXcgUm9tYW4iPml0IG9ubHkgaGFzIGZvcm1hdCBvZiB0d28gZGltZW50 aW9ucyBiZWluZyANCnJlYWQgYnkgeGdyYXBoLDwvRk9OVD48L0RJVj4NCjxESVY+PEZPTlQgZmFj ZT0iVGltZXMgTmV3IFJvbWFuIj5idXQgSSBkb24ndCBrbm93IGhvdyB0byBnZW5lcmF0ZSA8L0ZP TlQ+PEZPTlQgDQpmYWNlPSJUaW1lcyBOZXcgUm9tYW4iPnRoZSBmb3JtYXQgZGF0YSw8L0ZPTlQ+ PC9ESVY+DQo8RElWPjxGT05UIGZhY2U9IlRpbWVzIE5ldyBSb21hbiI+YW55IGhlbHAgaXMgYXBw cmVjaWF0ZWQuPC9GT05UPjwvRElWPg0KPERJVj4mbmJzcDs8L0RJVj4NCjxESVY+PEZPTlQgZmFj ZT0iVGltZXMgTmV3IFJvbWFuIj4mbmJzcDsmbmJzcDsmbmJzcDsgJm5ic3A7Jm5ic3A7Jm5ic3A7 IA0KJm5ic3A7Jm5ic3A7Jm5ic3A7ICZuYnNwOyZuYnNwOyZuYnNwOyAmbmJzcDsmbmJzcDsmbmJz cDsgSmltIA0KWWFuZzwvRk9OVD48L0RJVj48L0JPRFk+PC9IVE1MPg0K ------=_NextPart_000_0014_01BF55DF.681C0D00-- From zl@imail.com.cn Mon Jan 3 08:49:41 2000 From: zl@imail.com.cn (zl@imail.com.cn) Date: Mon, 3 Jan 2000 16:49:41 +0800 (CST) Subject: How to get separate data on flowmon? Message-ID: Dear Sir: I am a newcomer using ns2.1b5. Recently, I try to simulate a few of flows on IP network. I can not get separate statistical data on each flow.(Instead, I get aggregate result.) All the disscusion from mailinglist tell me that I must make use of "awk" to abstract the result from large files. I only need separate statistical data.Does anyone know the easy solution? Thanks. Zhang Lei ---------------------------------------------- 欢迎您使用 百家商务电子邮件系统 http://www.email.com.cn Welcome to E-mail business system From Fusun Inanc Mon Jan 3 22:28:08 2000 From: Fusun Inanc (Fusun Inanc) Date: Mon, 3 Jan 2000 17:28:08 -0500 (EST) Subject: codeword.cc in ns... Message-ID: <200001032228.RAA15663@poptop.> Hello everybody, I have a question. I want to use codeword.cc code in my simulation for my project. What I really want to do is generate data encode it using my code. After a lossy link I want to decode it and measure the throughput and goodput. For this reason I need to use codeword.cc code and also I will modify it. But at first I need to know how to use it. I figured out that mftp.cc uses the codeword.cc but I couldn't figured out how. If someone can help me on this I appreciate it. Thanks in advance.. Happy new year.. Fusun From salehi@isi.edu Tue Jan 4 01:12:45 2000 From: salehi@isi.edu (Nader Salehi) Date: Mon, 3 Jan 2000 17:12:45 -0800 (PST) Subject: codeword.cc in ns... In-Reply-To: <200001032228.RAA15663@poptop.> References: <200001032228.RAA15663@poptop.> Message-ID: <200001040112.RAA01717@joy.isi.edu> Fusun, codeword.cc is a contributed code and is not supported by the NS development team. Please contact Christoph Haenle at chris@cs.vu.nl. Nader >>> "Fusun" == Fusun Inanc writes: Fusun> Hello everybody, Fusun> I have a question. I want to use codeword.cc code in my Fusun> simulation for my project. Fusun> What I really want to do is generate data encode it using my Fusun> code. After a lossy link I want to decode it and measure the Fusun> throughput and goodput. Fusun> For this reason I need to use codeword.cc code and also I will Fusun> modify it. But at first I need to know how to use it. I Fusun> figured out that mftp.cc uses the codeword.cc but I couldn't Fusun> figured out how. Fusun> If someone can help me on this I appreciate it. Thanks in Fusun> advance.. Fusun> Happy new year.. Fusun> Fusun From b5703024@cc.ntu.edu.tw Tue Jan 4 08:06:58 2000 From: b5703024@cc.ntu.edu.tw (蕂) Date: Tue, 4 Jan 2000 16:06:58 +0800 Subject: ELN and snoop Message-ID: <002b01bf568a$c35d2400$8d14708c@r7942050> hello Recently , I want to get the code about "ELN" and "snoop" Can you tell me where to get it? thank you. wbliu From engp9374@nus.edu.sg Tue Jan 4 08:38:39 2000 From: engp9374@nus.edu.sg (Ge Yu) Date: Tue, 4 Jan 2000 16:38:39 +0800 Subject: SIP simulation Message-ID: <46AA89AF2D93D3118B0E00805F319E77EB852F@exs23.ex.nus.edu.sg> hi, friends! I'm working with SIP simulation in NS, is there anybody else who is doing or has touched the same thing? thank you & best regards Ge Yu From aibrahim@enst.fr Tue Jan 4 12:07:23 2000 From: aibrahim@enst.fr (abbas ibrahim) Date: Tue, 04 Jan 2000 12:07:23 +0000 Subject: question Message-ID: <3871E27B.59486CB1@enst.fr> good morning. i need to implement a special MAC protocol using CDMA. i hope you can inform me if there is an analegeous one implemented and where i can find it. in fact i need to implement the notion of frame then of codes. thank you very much. abbas From Tarik Alj Tue Jan 4 18:41:47 2000 From: Tarik Alj (Tarik Alj) Date: Tue, 4 Jan 2000 13:41:47 -0500 (EST) Subject: question Message-ID: <200001041841.NAA10071@valence.INRS-Telecom.UQuebec.CA> There is code for CSMA/CD/CA, however I don't know if it is any good. All files concerning MAC can be found in the ns directory: grep for mac. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Tue, 04 Jan 2000 12:07:23 +0000 > From: abbas ibrahim > X-Accept-Language: en > MIME-Version: 1.0 > To: ns-users@mash.cs.berkeley.edu > Subject: question > Content-Transfer-Encoding: 7bit > Content-Transfer-Encoding: 7bit > > good morning. > i need to implement a special MAC protocol using CDMA. i hope you can > inform me if there is an analegeous one implemented and where i can find > it. > in fact i need to implement the notion of frame then of codes. > thank you very much. > abbas Tarik From hungw@jupiter.nal.utoronto.ca Wed Jan 5 00:29:01 2000 From: hungw@jupiter.nal.utoronto.ca (Wing-Chung Hung) Date: Tue, 4 Jan 2000 19:29:01 -0500 (EST) Subject: vlan.tcl delay - bug? Message-ID: Hi, In vlan.tcl of ns-2.1b6-current, link delay is set to the delay of LL. However, I think the delay should be set to that of channel becasue channel is when the delay really occured. Can someone comment? Do I missing or misunderstand something? Thanks. -- Wing-Chung from vlan.tcl: LanNode instproc addNode {nodes bw delay {llType ""} {ifqType ""} \ {macType ""} {phyType ""}} { ... set ll [$nif set ll_] $ll set delay_ $delay $ll varp $varp_ } there is no setting of channel_::delay_ in vlan.tcl From zl@imail.com.cn Wed Jan 5 03:02:41 2000 From: zl@imail.com.cn (zl@imail.com.cn) Date: Wed, 5 Jan 2000 11:02:41 +0800 (CST) Subject: Application data constant:210 ?! Message-ID: Dear Sir: I am a newcomer using ns2.1b5. Recently, I try to simulate a few of flows on IP network. I set the application data via the instance of OTCL class Application/Traffic/CBR Application/Traffic/Exponential I want to change the parameter "packetSize_" as I wish. But I find the result is strange. The application packet always stays at 210 bytes each no matter what value I set. Is this a bug? Or,Maybe I forget to set some value. Thanks. Zhang Lei ---------------------------------------------- 欢迎您使用 百家商务电子邮件系统 http://www.email.com.cn Welcome to E-mail business system From zl@imail.com.cn Wed Jan 5 09:19:58 2000 From: zl@imail.com.cn (zl@imail.com.cn) Date: Wed, 5 Jan 2000 17:19:58 +0800 (CST) Subject: Material on RTP agent Message-ID: Dear sir: I have read the book "everyting.ps" throughly and searched your site for material on RTP agent. The book told me much about TCP and UDP,but little about RTP agent. Could you tell me about the detail meaning of parameters about Agent/RTP and its subclass? Thanks a lot Zhang Lei ---------------------------------------------- 欢迎您使用 百家商务电子邮件系统 http://www.email.com.cn Welcome to E-mail business system From barenco@dit.upm.es Wed Jan 5 09:52:59 2000 From: barenco@dit.upm.es (Claudia Jacy Barenco) Date: Wed, 5 Jan 2000 10:52:59 +0100 Subject: Simulation of MAC/Time Slot Message-ID: <001c01bf5762$a803dc20$67d5523e@fofinha> Dear users, I would like to simulate a MAC based on time slot (TDMA). Is there any code ready ? Thanks for attention. **************************************** Cl醬dia Jacy Barenco Polythecnic University of Madrid Telematic Systems - DIT Madrid - Spain ***************************************** From aibrahim@enst.fr Wed Jan 5 11:12:27 2000 From: aibrahim@enst.fr (abbas ibrahim) Date: Wed, 05 Jan 2000 11:12:27 +0000 Subject: question References: <200001041841.NAA10071@valence.INRS-Telecom.UQuebec.CA> Message-ID: <3873271A.D3193DF4@enst.fr> hi. i do not have the directory you mention. i have dowloaded my own ns from berkeley site and i think it is a recent one. if you have this directory, please send it to me. thank you abbas Tarik Alj wrote: > There is code for CSMA/CD/CA, however I don't know if it is any good. All files > concerning MAC can be found in the ns directory: grep for mac. > > > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to > owner-ns-users@irtf.CS.Berkeley.EDU using -f > > Date: Tue, 04 Jan 2000 12:07:23 +0000 > > From: abbas ibrahim > > X-Accept-Language: en > > MIME-Version: 1.0 > > To: ns-users@mash.cs.berkeley.edu > > Subject: question > > Content-Transfer-Encoding: 7bit > > Content-Transfer-Encoding: 7bit > > > > good morning. > > i need to implement a special MAC protocol using CDMA. i hope you can > > inform me if there is an analegeous one implemented and where i can find > > it. > > in fact i need to implement the notion of frame then of codes. > > thank you very much. > > abbas > > Tarik From orlandini.federico@iol.it Wed Jan 5 10:47:12 2000 From: orlandini.federico@iol.it (Federico Orlandini) Date: Wed, 5 Jan 2000 11:47:12 +0100 Subject: Application/Traffic/CBR bug ? Message-ID: <001101bf576a$3c132220$1e440d97@hal9000> Hi all, I need to simulate a VBR source on UDP with changes its sending rate every 1/24 sec. ( 24 fps ). To do it i'm using a CBR traffic generator and i change its sending rate ( rate_ variable ) every 1/24 second. The new sending rate is a random variable generated by a RandomVariable/Exponential object. Sometimes the CBR generator seems not to work because it stops itself and if i change its sending rate it seems do nothing. To see what i'm trying to say please take a look to the following Tcl-script. set ns [new Simulator] set file1 [open /home/federico/ns-result/out.nam w] set filetrace [open /home/federico/ns-result/delme.tr w] set cbrfile [open /home/federico/ns-result/cbr.tr w] ; # This file is used to collect CBR Source's data $ns namtrace-all $file1 set NumBytePrec1 0 set TimePrec1 0 set node1 [$ns node] set node2 [$ns node] $ns simplex-link $node1 $node2 110Mb 10ms DropTail set agent1 [new Agent/UDP] $ns attach-agent $node1 $agent1 set traffic [new Application/Traffic/CBR] $traffic attach-agent $agent1 set agent2 [new Agent/Null] $ns attach-agent $node2 $agent2 $ns connect $agent1 $agent2 set AvgRate 3000000 proc finish {} { global ns file1 filetrace cbrfile $ns flush-trace close $cbrfile close $filetrace close $file1 exec /home/federico/ns-allinone-2.1b5/nam-1.0a7/nam /home/federico/ns-result/out.nam & exec /home/federico/ns-allinone-2.1b5/xgraph/xgraph /home/federico/ns-result/delme.tr -geometry 400x400 & exit 1 } ############################################################################ ########### # This procedure inserts a snooper between node1 and node2 to read the number of packets sending from CBR# ############################################################################ ########### SimpleLink instproc attach-snooper-in { SrcNode DstNode } { global ns $self instvar Snooper Entry Link TtlPointer Monitor # Create a SnoopQueue/In object set Snooper [new SnoopQueue/In] puts stdout "Create Snooper $Snooper" # Get reference to the entry point for DstNode set Entry [$DstNode entry] # Get reference to the ttl block set Link [$ns link $SrcNode $DstNode] set TtlPointer [$Link set ttl_] #Insert the snooper between ttl and entry-point $lPointer target $Snooper $Snooper target $Entry # Creating a Queuemonitor to collect data set Monitor [new QueueMonitor] puts "Create Monitor : $Monitor" $Snooper set-monitor $Monitor return $Monitor } ########################################################################## # This procedure read statistics and write it in a file # ########################################################################## proc ReadMonitorData { File Fileb Monitor Stop Step } { global ns NumBytePrec1 TimePrec1 # Leggiamo il tempo set now [$ns now] # Leggiamo il numero di bytes arrivati set Bytes [$Monitor set barrivals_] # Controlliamo quali dati dobbiamo processare if {$File != "NULL"} then { # Scriviamo i dati nel file d'uscita puts $File "$now $Bytes" # puts stdout "Monitor Data : $Bytes" Utile per il Debug } if {$Fileb != "NULL"} then { set DeltaByte [expr $Bytes-$NumBytePrec1] set DeltaTime [expr $now-$TimePrec1] puts $Fileb "$now [expr $DeltaByte/$DeltaTime]" # puts stdout "Monitor Band : [expr $DeltaByte/$DeltaTime]" Per il Debug } if {$File == "NULL" && $FIleb == "NULL"} then { puts stdout "Error : procedure ReadMonitorData can't collect data." puts stdout " Both args are NULL." exit 1 ; } # Rischeduliamo il processo solo se non eccede la fine della simulazione set Next [expr $now+$Step] if { $Next > $Stop } then { return } $ns at $Next "ReadMonitorData $File $Fileb $Monitor $Stop $Step" # Aggiorniamo i dati NumBytePrec1 e TimePrec1 set $NumBytePrec1 $Bytes set $TimePrec1 $now } Application/Traffic/CBR instproc Prepara {} { global ns AvgRate $self instvar rate_ $self set Rng [new RNG] $self instvar Rng $Rng seed 1 $self set RndGen [new RandomVariable/Exponential] $self instvar RndGen $RndGen use-rng $Rng $RndGen set avg_ $AvgRate } set link [$ns link $node1 $node2] set monitor [$link attach-snooper-in $node1 $node2] $ns at 0.0 "$traffic start" $ns at 0.0 "$traffic Prepara" $ns at 0.0 "ReadMonitorData $filetrace NULL $monitor 100.0 0.01" $ns at 0.0 "$traffic set rate_ 1Mb" $ns at 5.0 "$traffic set rate_ 10000" ; # <--- Try to change this rate , for example 10 , 100 , 1000 ,..... $ns at 6.0 "$traffic set rate_ 1Mb" $ns at 100.0 "$traffic stop" $ns at 100.1 "finish" $ns run If you see the graph generated you can see that at 6.0 the rate not change although we command it to change the cbr sending rate. Do you think it is a bug or i'm wronging something ? Thank you and have a happy new year. Federico Orlandini E-Mail : orlandini.federico@iol.it From guru0109@palgong.knu.ac.kr Wed Jan 5 13:24:48 2000 From: guru0109@palgong.knu.ac.kr (=?euc-kr?B?w9a/tbz2?=) Date: Wed, 5 Jan 2000 22:24:48 +0900 Subject: [Q] SCFQ or SFQ implementation available? Message-ID: <000701bf5780$3d2db7e0$d50ee69b@knu.ac.kr> Hi all, I found WFQ, CBQ, DRR... some scheduling code in NS site. But can't find SCFQ(Selfclocked Fair Queueing) or SFQ(Start Time Fair Queueing). Does anybody implement these scheduling? Any information fine. Thx in advance. ______________________________________________________ 弥康荐 (ChoiYoungSoo) e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 TeNet Lab : http://tenet.knu.ac.kr/ From Tarik Alj Wed Jan 5 15:44:21 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 5 Jan 2000 10:44:21 -0500 (EST) Subject: vlan.tcl delay - bug? Message-ID: <200001051544.KAA11134@valence.INRS-Telecom.UQuebec.CA> that delay is not a transmission delay, but more of a latency delay; as I understand it... > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Tue, 4 Jan 2000 19:29:01 -0500 (EST) > From: Wing-Chung Hung > X-Sender: hungw@uranus > To: ns-users@mash.cs.berkeley.edu > Subject: vlan.tcl delay - bug? > MIME-Version: 1.0 > > > Hi, > > In vlan.tcl of ns-2.1b6-current, link delay is set to the delay of LL. > However, I think the delay should be set to that of channel becasue > channel is when the delay really occured. > > Can someone comment? Do I missing or misunderstand something? > > Thanks. > > -- Wing-Chung > > > from vlan.tcl: > > LanNode instproc addNode {nodes bw delay {llType ""} {ifqType ""} \ > {macType ""} {phyType ""}} { > > ... > > set ll [$nif set ll_] > $ll set delay_ $delay > $ll varp $varp_ > } > > there is no setting of channel_::delay_ in vlan.tcl Tarik From Fusun Inanc Wed Jan 5 16:20:21 2000 From: Fusun Inanc (Fusun Inanc) Date: Wed, 5 Jan 2000 11:20:21 -0500 (EST) Subject: bits vs. packets Message-ID: <200001051620.LAA18348@poptop.> Dear ns users, In my simulation, I want to send bits instead of packets. Is it possible? If someone help me I would appreciate it. Thanks in advance Fusun Inanc From L.Wood@eim.surrey.ac.uk Wed Jan 5 17:01:50 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 5 Jan 2000 17:01:50 +0000 (GMT) Subject: cvsweb problem Message-ID: The ns/mash cvs webtree is currently non-functional - as a result of upgrading to Apache 1.3.9 from 1.2b7? please take a look - I'm rather reliant on this for keeping on top of ns. thanks, L. PGP http://www-mash.cs.berkeley.edu/cgi-bin/cvsweb Internal Server Error The server encountered an internal error or misconfiguration and was unable to complete your request. Please contact the server administrator, webmaster@mash.CS.Berkeley.EDU and inform them of the time the error occurred, and anything you might have done that may have caused the error. More information about this error may be available in the server error log. Apache/1.3.9 Server at www-mash.CS.Berkeley.EDU Port 80 From hungw@jupiter.nal.utoronto.ca Wed Jan 5 19:51:00 2000 From: hungw@jupiter.nal.utoronto.ca (Wing-Chung Hung) Date: Wed, 5 Jan 2000 14:51:00 -0500 (EST) Subject: vlan.tcl delay - bug? In-Reply-To: <200001051544.KAA11134@valence.INRS-Telecom.UQuebec.CA> Message-ID: Here are my understand: 1. LL is a child of DelayLink which is a child of link. In a point to point connection (not lan), the delay_ of link, I think, is the propagation delay (the time it take for a bit of information from transmitter to receiver) and there is no channel in such topology. 2. when I create a lan using make-lan set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] I am thinking that the $delay is the delay is the propagation (channel) delay between a transmitting and a receiving node. I.e. it should be the delay_ of a channel. In channel.cc, delay_ actually means propagation dealy and is used for collison calculation. double Channel::get_pdelay(Node* /*tnode*/, Node* /*rnode*/) { // Dummy function return delay_; } By assigning the $delay to the delay_ of LL, this delay means the amount of time the packet stay in a node before being send to it's downtarget_ which is the queue. This is reflected in ll.cc from ll.cc void LL::sendDown(Packet* p) { ... if (tx == 0) { Scheduler& s = Scheduler::instance(); s.schedule(downtarget_, p, delay_); } } My conclusion is that the $delay argument in make-lan set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] means the inbound processing delay (the amount of time to stay in the node before being queue for transmission), and is not related to propagation dealy. Please comment. Thanks. -- Wing-Chung On Wed, 5 Jan 2000, Tarik Alj wrote: > that delay is not a transmission delay, but more of a latency delay; as I > understand it... > > > > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to > owner-ns-users@irtf.CS.Berkeley.EDU using -f > > Date: Tue, 4 Jan 2000 19:29:01 -0500 (EST) > > From: Wing-Chung Hung > > X-Sender: hungw@uranus > > To: ns-users@mash.cs.berkeley.edu > > Subject: vlan.tcl delay - bug? > > MIME-Version: 1.0 > > > > > > Hi, > > > > In vlan.tcl of ns-2.1b6-current, link delay is set to the delay of LL. > > However, I think the delay should be set to that of channel becasue > > channel is when the delay really occured. > > > > Can someone comment? Do I missing or misunderstand something? > > > > Thanks. > > > > -- Wing-Chung > > > > > > from vlan.tcl: > > > > LanNode instproc addNode {nodes bw delay {llType ""} {ifqType ""} \ > > {macType ""} {phyType ""}} { > > > > ... > > > > set ll [$nif set ll_] > > $ll set delay_ $delay > > $ll varp $varp_ > > } > > > > there is no setting of channel_::delay_ in vlan.tcl > > Tarik > > From Darren.Kressaty@sbirs.specastro.com Wed Jan 5 21:42:11 2000 From: Darren.Kressaty@sbirs.specastro.com (Kressaty, Darren (EDI@SA)) Date: Wed, 5 Jan 2000 14:42:11 -0700 Subject: Bonehead Question... Message-ID: <1D0BA26DBE71D3118C7100508B5B0B1E0581C9@AZEXCHANGE01> Ok, I know this is a bone-head, FNG question. But I have been wrestling with this thing all day. I just installed the latest version of Red Hat, and the latest ns-allinone... The files are all in the right place ( /usr/local/NS/..... ). When I'm trying to execute the ./install file from the 'allinone' directory I get the 'file not found/does not exist' error. I have added the directory to my PATH statement in my .cshrc file, moved it's location to the front of the PATH statement, checked permissions on the files and directories. Everything looks to be in order. What am I doing wrong, or what should I try that I already haven't? Thanks Darren Kressaty From yanyu@catarina.usc.edu Wed Jan 5 22:59:39 2000 From: yanyu@catarina.usc.edu (Yan Yu) Date: Wed, 5 Jan 2000 14:59:39 -0800 (PST) Subject: CVS Message-ID: Hello, could someone pls check why read-only CVS history on ns homepage does not work? Thanks in advance! yan From johnh@isi.edu Thu Jan 6 00:36:49 2000 From: johnh@isi.edu (John Heidemann) Date: Wed, 05 Jan 2000 16:36:49 -0800 Subject: CVS In-Reply-To: Message from Yan Yu of "Wed, 05 Jan 2000 14:59:39 PST." Message-ID: <200001060036.QAA22222@dash.isi.edu> >could someone pls check why read-only CVS history on ns homepage does not >work? >Thanks in advance! To Yan and Lloyd and the others who have asked about CVSweb on ns: Mash was upgraded (from an ancient to a modern version of FreeBSD) mid-December. We're trying to get some of the things that broken straightened out, but it may be a few days. -John Heidemann From salehi@isi.edu Thu Jan 6 03:03:37 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 5 Jan 2000 19:03:37 -0800 (PST) Subject: Bonehead Question... In-Reply-To: <1D0BA26DBE71D3118C7100508B5B0B1E0581C9@AZEXCHANGE01> References: <1D0BA26DBE71D3118C7100508B5B0B1E0581C9@AZEXCHANGE01> Message-ID: <200001060303.TAA14917@joy.isi.edu> Re-run install and dump the output into a file and send it to me. Nader >>> "EDI@SA" == EDI@SA writes: EDI@SA> Ok, I know this is a bone-head, FNG question. But I have been EDI@SA> wrestling with this thing all day. EDI@SA> I just installed the latest version of Red Hat, and the latest EDI@SA> ns-allinone... The files are all in the right place ( EDI@SA> /usr/local/NS/..... ). When I'm trying to execute the EDI@SA> ./install file from the 'allinone' directory I get the 'file EDI@SA> not found/does not exist' error. I have added the directory to EDI@SA> my PATH statement in my .cshrc file, moved it's location to EDI@SA> the front of the PATH statement, checked permissions on the EDI@SA> files and directories. Everything looks to be in order. What EDI@SA> am I doing wrong, or what should I try that I already haven't? EDI@SA> Thanks Darren Kressaty From salehi@isi.edu Thu Jan 6 03:28:15 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 5 Jan 2000 19:28:15 -0800 (PST) Subject: bits vs. packets In-Reply-To: <200001051620.LAA18348@poptop.> References: <200001051620.LAA18348@poptop.> Message-ID: <200001060328.TAA14988@joy.isi.edu> No It's not possible. Nader >>> "Fusun" == Fusun Inanc writes: Fusun> Dear ns users, In my simulation, I want to send bits instead of Fusun> packets. Is it possible? If someone help me I would appreciate Fusun> it. Fusun> Thanks in advance Fusun> Fusun Inanc From drlab@sds.shtdu.edu.cn Thu Jan 6 04:57:15 2000 From: drlab@sds.shtdu.edu.cn (wang xu) Date: Thu, 6 Jan 2000 12:57:15 +0800 Subject: Application/Traffic/CBR bug ? In-Reply-To: <001101bf576a$3c132220$1e440d97@hal9000> References: <001101bf576a$3c132220$1e440d97@hal9000> Message-ID: <00010613074800.00635@s168p19.shtdu.edu.cn> Hi Federico, I think the the graph generated is reasonable. at 5.0 the rate change to 10000 and at 6.0 it change to 1M again follow your command. $ns at 6.0 "$traffic set rate_ 1Mb" I changed your code at line : $ns at 6.0 "$traffic set rate_ 1Mb" to $ns at 6.0 "$traffic set rate_ 2Mb" and run ns again then the output graph is noticeble sharp at 6.0. did I misunderstand your meaning? Regards. On Wed, 05 Jan 2000, you wrote: > > set link [$ns link $node1 $node2] > set monitor [$link attach-snooper-in $node1 $node2] > > $ns at 0.0 "$traffic start" > $ns at 0.0 "$traffic Prepara" > $ns at 0.0 "ReadMonitorData $filetrace NULL $monitor 100.0 0.01" > $ns at 0.0 "$traffic set rate_ 1Mb" > $ns at 5.0 "$traffic set rate_ 10000" ; # <--- Try to change this rate , for > example 10 , 100 , 1000 ,..... > $ns at 6.0 "$traffic set rate_ 1Mb" > $ns at 100.0 "$traffic stop" > $ns at 100.1 "finish" > > $ns run > > If you see the graph generated you can see that at 6.0 the rate not change > although we command it to change the cbr sending rate. > > Do you think it is a bug or i'm wronging something ? > Thank you and have a happy new year. > > > Federico Orlandini > E-Mail : orlandini.federico@iol.it -- Wang Xu Email:drlab@shtdu.edu.cn From lhyang@cti.com.cn Thu Jan 6 06:08:16 2000 From: lhyang@cti.com.cn (yanglihai) Date: Thu, 6 Jan 2000 14:08:16 +0800 Subject: questions of static class? Message-ID: <000a01bf580c$6c8b8d00$350279c8@jimyang> This is a multi-part message in MIME format. ------=_NextPart_000_0007_01BF584F.7A288600 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 aGkgYWxsLA0KICAgIHdobyBjYW4gdGVsbCBtZSB3aGF0IGFjdGlvbiBkb2VzIHRoZSBzdGF0aWMg Y2xhc3M/DQpob3cgaXQgYWN0IGFzID8NCmUuZy4NCg0Kc3RhdGljIGNsYXNzIERyb3BUYWlsQ2xh c3MgOiBwdWJsaWMgVGNsQ2xhc3Mgew0KIHB1YmxpYzoNCiBEcm9wVGFpbENsYXNzKCkgOiBUY2xD bGFzcygiUXVldWUvRHJvcFRhaWwiKSB7fQ0KIFRjbE9iamVjdCogY3JlYXRlKGludCwgY29uc3Qg Y2hhcipjb25zdCopIHsNCiAgcmV0dXJuIChuZXcgRHJvcFRhaWwpOw0KIH0NCn0gY2xhc3NfZHJv cF90YWlsOw0KDQphbnkgaGVscCBpcyBhcHByZWNpYXRlZC4NCg0KICAgICAgICAgICAgICAgIEpp bSBZYW5nDQoNCg0KDQo= ------=_NextPart_000_0007_01BF584F.7A288600 Content-Type: text/html; charset="gb2312" Content-Transfer-Encoding: base64 PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWdi MjMxMiIgaHR0cC1lcXVpdj1Db250ZW50LVR5cGU+DQo8TUVUQSBjb250ZW50PSJNU0hUTUwgNS4w MC4yMDE0LjIxMCIgbmFtZT1HRU5FUkFUT1I+DQo8U1RZTEU+PC9TVFlMRT4NCjwvSEVBRD4NCjxC T0RZIGJnQ29sb3I9I2ZmZmZmZj4NCjxESVY+PEZPTlQgZmFjZT0iVGltZXMgTmV3IFJvbWFuIj5o aSBhbGwsPC9GT05UPjwvRElWPg0KPERJVj48Rk9OVCBmYWNlPSJUaW1lcyBOZXcgUm9tYW4iPiZu YnNwOyZuYnNwOyZuYnNwOyB3aG8gY2FuIHRlbGwgbWUgd2hhdCANCmFjdGlvbiZuYnNwO2RvZXMg dGhlIHN0YXRpYyBjbGFzcz88L0ZPTlQ+PC9ESVY+DQo8RElWPjxGT05UIGZhY2U9IlRpbWVzIE5l dyBSb21hbiI+aG93IGl0IGFjdCBhcyA/PC9GT05UPjwvRElWPg0KPERJVj48Rk9OVCBzaXplPTI+ ZS5nLjwvRk9OVD48L0RJVj4NCjxESVY+Jm5ic3A7PC9ESVY+DQo8RElWPjxGT05UIGZhY2U9IlRp bWVzIE5ldyBSb21hbiI+c3RhdGljIGNsYXNzIERyb3BUYWlsQ2xhc3MgOiBwdWJsaWMgVGNsQ2xh c3MgDQp7PEJSPiZuYnNwO3B1YmxpYzo8QlI+Jm5ic3A7RHJvcFRhaWxDbGFzcygpIDogVGNsQ2xh c3MoIlF1ZXVlL0Ryb3BUYWlsIikgDQp7fTxCUj4mbmJzcDtUY2xPYmplY3QqIGNyZWF0ZShpbnQs IGNvbnN0IGNoYXIqY29uc3QqKSB7PEJSPiZuYnNwOyZuYnNwO3JldHVybiANCihuZXcgRHJvcFRh aWwpOzxCUj4mbmJzcDt9PEJSPn0gY2xhc3NfZHJvcF90YWlsOzwvRk9OVD48L0RJVj4NCjxESVY+ Jm5ic3A7PC9ESVY+DQo8RElWPjxGT05UIGZhY2U9IlRpbWVzIE5ldyBSb21hbiI+YW55IGhlbHAg aXMgYXBwcmVjaWF0ZWQuPC9GT05UPjwvRElWPg0KPERJVj4mbmJzcDs8L0RJVj4NCjxESVY+PEZP TlQgZmFjZT0iVGltZXMgTmV3IFJvbWFuIj4mbmJzcDsmbmJzcDsmbmJzcDsgJm5ic3A7Jm5ic3A7 Jm5ic3A7IA0KJm5ic3A7Jm5ic3A7Jm5ic3A7ICZuYnNwOyZuYnNwOyZuYnNwOyBKaW0gWWFuZzwv Rk9OVD48L0RJVj4NCjxESVY+Jm5ic3A7PC9ESVY+DQo8RElWPiZuYnNwOzwvRElWPg0KPERJVj4m bmJzcDs8L0RJVj48L0JPRFk+PC9IVE1MPg0K ------=_NextPart_000_0007_01BF584F.7A288600-- From Hajer.Tounsi@ensi.rnu.tn Thu Jan 6 09:42:24 2000 From: Hajer.Tounsi@ensi.rnu.tn (Hajer Tounsi) Date: Thu, 06 Jan 2000 10:42:24 +0100 Subject: new classifier? Message-ID: <38746380.89997C8B@ensi.rnu.tn> Hi, I have defined a new classifier (Classifier/IPAgrege), I have added in Makefile and Makefile.in the file.o referencing the corresponding Tclclass and methods, but when I use [new Classifier/IPAgrege] in my simulation, I get ..... invalid command Classifier/IPAgrege Everything looks to be in order. What am I doing wrong, or what should I try that I already haven't? Thanks Hajer TOUNSI From guru0109@palgong.knu.ac.kr Thu Jan 6 13:45:21 2000 From: guru0109@palgong.knu.ac.kr (=?euc-kr?B?w9a/tbz2?=) Date: Thu, 6 Jan 2000 22:45:21 +0900 Subject: How can I get *Packet???? Message-ID: <002901bf584c$469fcda0$d50ee69b@knu.ac.kr> Hi all.... I'm modifying DRR. And run NS, NS.core occure.. Here is my problem. Original DRR deque operation... ----snip--------- pkt=curr->lookup(0); ch=(hdr_cmn*)pkt->access(off_cmn_); iph= (hdr_ip*)pkt->access(off_ip_); if (curr->deficitCounter >= ch->size()) { curr->deficitCounter -= ch->size(); pkt=curr->deque(); curr->bcount -= ch->size(); --curr->pkts; ----snip--------- And here is mine ----- snip---------- while (!pkt) { ... cout << "Curr Pointer" << curr << endl; cout << "Curr pkts" << curr->pkts << endl; ... pkt=curr->lookup(0); ch=(hdr_cmn*)pkt->access(off_cmn_); iph= (hdr_ip*)pkt->access(off_ip_); pkt=curr->deque(); curr->bcount -= ch->size(); --curr->pkts; --pktcnt; bytecnt -= ch->size(); .... cout << "Curr pkts" << curr->pkts << endl; if (curr->pkts == 0) { curr->turn=0; --flwcnt; curr->deficitCounter=0; curr=curr->idle(curr); } else { pkt=curr->lookup(0); // HERE ERROR! CORE DUMP cout << "pkt pointer" << pkt << endl; ch=(hdr_cmn*)pkt->access(off_cmn_); iph=(hdr_ip*)pkt->access(off_ip_); .... While output, first current pkts(packet counter) = 7 and after deque, curr->pkts = 6. so it seems that first packet served and remove frome the curr queue. but when i access again using lookup(0) or lookup(1), (to access 2nd packet) NS comre dumped... I want that service 1st packet and read 2nd packet's header information(not serving) How can i access 2nd packet's header? Did I miss something? Any infomation fine. Thx in advance. ______________________________________________________ ChoiYoungSoo e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 From ppieda@nortelnetworks.com Thu Jan 6 15:16:11 2000 From: ppieda@nortelnetworks.com (Peter Pieda) Date: Thu, 6 Jan 2000 09:16:11 -0600 Subject: Monitoring duration of a Tcp Flow Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF5858.F7487330 Content-Type: text/plain Hello, Is it possible to measure the duration of a Tcp Flow? I can get the Flow start time but I am having difficulty finding when the Flow ends. I looked at using a Flowmon monitor but it tracks qualities in bytes or packets. I'm working with the PagePool/WebTraf model. I am trying to determine the length of each transaction (or sending time for one object) during the course of a Web session. Thanks Peter ppieda@nortelnetworks.com ------_=_NextPart_001_01BF5858.F7487330 Content-Type: text/html Content-Transfer-Encoding: quoted-printable Monitoring duration of a Tcp Flow

Hello,

Is it possible to = measure the duration of a Tcp Flow?

I can get the Flow = start time but I am having difficulty finding when the Flow
ends. I looked at = using a Flowmon monitor but it tracks qualities in bytes or = packets.

I'm working with the = PagePool/WebTraf model. I am trying to determine
the length of each = transaction (or sending time for one object) during the course
of a Web = session.

Thanks

Peter
ppieda@nortelnetworks.com

------_=_NextPart_001_01BF5858.F7487330-- From riley@cc.gatech.edu Thu Jan 6 15:24:48 2000 From: riley@cc.gatech.edu (George Riley) Date: Thu, 6 Jan 2000 10:24:48 -0500 (EST) Subject: Monitoring duration of a Tcp Flow In-Reply-To: Message-ID: You can define a "done" proc for TCP agents that will be called when there is no more data to send on any connection. See below... Agent/TCP instproc done {} { global ns donecount $self instvar node_ dst_ set peerid [expr $dst_ >> [Simulator set NodeShift_]] puts "Node [$node_ id], peer $peerid, all done at [$ns now]" incr donecount ... whatever else you need } George -George F. Riley (riley@cc.gatech.edu) On Thu, 6 Jan 2000, Peter Pieda wrote: > Hello, > > Is it possible to measure the duration of a Tcp Flow? > > I can get the Flow start time but I am having difficulty finding when the > Flow > ends. I looked at using a Flowmon monitor but it tracks qualities in bytes > or packets. > > I'm working with the PagePool/WebTraf model. I am trying to determine > the length of each transaction (or sending time for one object) during the > course > of a Web session. > > Thanks > > Peter > ppieda@nortelnetworks.com > From Tarik Alj Thu Jan 6 15:39:15 2000 From: Tarik Alj (Tarik Alj) Date: Thu, 6 Jan 2000 10:39:15 -0500 (EST) Subject: vlan.tcl delay - bug? Message-ID: <200001061539.KAA01680@opale.INRS-Telecom.UQuebec.CA> If you are using mac-802.3 you should be aware that collision detection is tightly linked to the maximum delay packets are experiencing (and therefore to the length of cable used and the bandwidth). LL is a child of linkdelay, not delaylink. linkdelay is really a class of connector that induces a delay, not some sort of link. vlan.tcl is "configured" for plain MAC, not mac-802.3. If you want to set the Channel delay, you can do so in LanIface::init. Now the Notes (now NS-Manual) say: Channel : delay_ is propagation delay on the Channel LL : delay_ is link-level overhead. So I would agree with your understanding. > Date: Wed, 5 Jan 2000 14:51:00 -0500 (EST) > From: Wing-Chung Hung > X-Sender: hungw@uranus > To: Tarik Alj > cc: ns users mail list > Subject: Re: vlan.tcl delay - bug? > MIME-Version: 1.0 > > > Here are my understand: > > 1. > > LL is a child of DelayLink which is a child of link. > In a point to point connection (not lan), the delay_ of link, I think, is > the propagation delay (the time it take for a bit of information from > transmitter to receiver) and there is no channel in such topology. > > 2. > > when I create a lan using make-lan > > set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] > > I am thinking that the $delay is the delay is the propagation > (channel) delay between a transmitting and a receiving node. I.e. it > should be the delay_ of a channel. > > In channel.cc, delay_ actually means propagation dealy and is used for > collison calculation. > > double > Channel::get_pdelay(Node* /*tnode*/, Node* /*rnode*/) > { > // Dummy function > return delay_; > } > > By assigning the $delay to the delay_ of LL, this delay means the amount > of time the packet stay in a node before being send to it's downtarget_ > which is the queue. This is reflected in ll.cc > > from ll.cc > > void LL::sendDown(Packet* p) > { > ... > if (tx == 0) { > Scheduler& s = Scheduler::instance(); > s.schedule(downtarget_, p, delay_); > } > } > > > My conclusion is that the $delay argument in make-lan > > set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] > > means the inbound processing delay (the amount of time to stay in the node > before being queue for transmission), and is not related to propagation > dealy. > > Please comment. > > Thanks. > > -- Wing-Chung > Tarik From leventi@fgan.de Thu Jan 6 15:39:49 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Thu, 6 Jan 2000 16:39:49 +0100 (MET) Subject: address-format Message-ID: <200001061539.QAA03552@melle.ffm.fgan.de> Dear ns users, I've been studying the ns Notes and Documentation, after I have installed the ns Packet at my Sun computer. Does anybody know some informations about the maximum value n (number of bits) reserved for the node addresses of the simulator? Is it so that only the value of 32 is allowed in the revised version? Has anyone made experience with something different? I would appreciate some hints A.Leventi-Peetz From Tarik Alj Thu Jan 6 16:09:02 2000 From: Tarik Alj (Tarik Alj) Date: Thu, 6 Jan 2000 11:09:02 -0500 (EST) Subject: vlan.tcl delay - bug? Message-ID: <200001061609.LAA01817@opale.INRS-Telecom.UQuebec.CA> > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Thu, 6 Jan 2000 10:39:15 -0500 (EST) > From: Tarik Alj > Subject: Re: vlan.tcl delay - bug? > To: hungw@jupiter.nal.utoronto.ca > Cc: ns-users@mash.cs.berkeley.edu > MIME-Version: 1.0 > Content-MD5: kw4zIt4OWoGiIcsNDhgfyg== > > If you are using mac-802.3 you should be aware that collision detection is > tightly linked to the maximum delay packets are experiencing (and therefore to > the length of cable used and the bandwidth). > > LL is a child of linkdelay, not delaylink. linkdelay is really a class of > connector that induces a delay, not some sort of link. > > vlan.tcl is "configured" for plain MAC, not mac-802.3. If you want to set the > Channel delay, you can do so in LanIface::init. Actually it's in LanNode::addNode > > Now the Notes (now NS-Manual) say: > > Channel : delay_ is propagation delay on the Channel > > LL : delay_ is link-level overhead. > > So I would agree with your understanding. > > > Date: Wed, 5 Jan 2000 14:51:00 -0500 (EST) > > From: Wing-Chung Hung > > X-Sender: hungw@uranus > > To: Tarik Alj > > cc: ns users mail list > > Subject: Re: vlan.tcl delay - bug? > > MIME-Version: 1.0 > > > > > > Here are my understand: > > > > 1. > > > > LL is a child of DelayLink which is a child of link. > > In a point to point connection (not lan), the delay_ of link, I think, is > > the propagation delay (the time it take for a bit of information from > > transmitter to receiver) and there is no channel in such topology. > > > > 2. > > > > when I create a lan using make-lan > > > > set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] > > > > I am thinking that the $delay is the delay is the propagation > > (channel) delay between a transmitting and a receiving node. I.e. it > > should be the delay_ of a channel. > > > > In channel.cc, delay_ actually means propagation dealy and is used for > > collison calculation. > > > > double > > Channel::get_pdelay(Node* /*tnode*/, Node* /*rnode*/) > > { > > // Dummy function > > return delay_; > > } > > > > By assigning the $delay to the delay_ of LL, this delay means the amount > > of time the packet stay in a node before being send to it's downtarget_ > > which is the queue. This is reflected in ll.cc > > > > from ll.cc > > > > void LL::sendDown(Packet* p) > > { > > ... > > if (tx == 0) { > > Scheduler& s = Scheduler::instance(); > > s.schedule(downtarget_, p, delay_); > > } > > } > > > > > > My conclusion is that the $delay argument in make-lan > > > > set lan [$ns make-lan $nodelist $bw $delay LL Queue/DropTail Mac] > > > > means the inbound processing delay (the amount of time to stay in the node > > before being queue for transmission), and is not related to propagation > > dealy. > > > > Please comment. > > > > Thanks. > > > > -- Wing-Chung > > > > Tarik Tarik From agandhi@cisunix.unh.edu Thu Jan 6 17:55:07 2000 From: agandhi@cisunix.unh.edu (Arun Gandhi) Date: Thu, 06 Jan 2000 12:55:07 -0500 Subject: Question Message-ID: <3874D6FB.B93027F8@cisunix.unh.edu> I am a new user. I am trying to get count of the packet drops. How can I get it. I am trying with qmon presently. Arun -- Arun Gandhi M.S. (Computer Science) Phone : 603-862-7614 (Res) University of New Hampshire, 603-862-3786 (Off) Durham, NH 03824 Email : agandhi@cisunix.unh.edu From johnh@isi.edu Thu Jan 6 21:48:40 2000 From: johnh@isi.edu (John Heidemann) Date: Thu, 06 Jan 2000 13:48:40 -0800 Subject: questions of static class? Message-ID: <200001062148.NAA28097@dash.isi.edu> In-Reply-To: Message from "yanglihai" of "Thu, 06 Jan 2000 14:08:16 +0800." <000a01bf580c$6c8b8d00$350279c8@jimyang> -------- >hi all, > who can tell me what action does the static class? >how it act as ? >e.g. > >static class DropTailClass : public TclClass { > public: > DropTailClass() : TclClass("Queue/DropTail") {} > TclObject* create(int, const char*const*) { > return (new DropTail); > } >} class_drop_tail; > >any help is appreciated. The construction of the class_drop_tail object (the one instance of this class) creates an agent of type Queue/DropTail in Tcl. See the code in TclCL for details, or the Tcl binding chapter of the manual. -John Heidemann From L.Wood@eim.surrey.ac.uk Thu Jan 6 22:35:55 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Thu, 6 Jan 2000 22:35:55 +0000 (GMT) Subject: CVSweb returns Message-ID: The ns/mash CVS tree: http://www-mash.cs.berkeley.edu/cgi-bin/cvsweb is back. thanks, brian. cheers, L. PGP ---------- Forwarded message ---------- Date: Thu, 06 Jan 2000 14:33:01 -0800 From: Brian K. Shiratsuki To: L.Wood@eim.surrey.ac.uk Cc: johnh@isi.edu Subject: Re: message resend - cvsweb 500 error -------- lloyd, please try cvsweb now; it should work better. -- thanks, brian From agandhi@cisunix.unh.edu Fri Jan 7 00:21:40 2000 From: agandhi@cisunix.unh.edu (Arun Gandhi) Date: Thu, 06 Jan 2000 19:21:40 -0500 Subject: Question Message-ID: <38753194.A77BC1E@cisunix.unh.edu> In the example I have I need to generate nodes after certain time that is after running simulation for some time.How can I dynamically generate a node.(source,intermediate or destination). I am trying to change topology dynamically.(at run time and see how is my network behaving) Regards, arun -- Arun Gandhi M.S. (Computer Science) Phone : 603-862-7614 (Res) University of New Hampshire, 603-862-3786 (Off) Durham, NH 03824 Email : agandhi@cisunix.unh.edu From engp9374@nus.edu.sg Fri Jan 7 02:13:14 2000 From: engp9374@nus.edu.sg (Ge Yu) Date: Fri, 7 Jan 2000 10:13:14 +0800 Subject: RSVP simulation Message-ID: <46AA89AF2D93D3118B0E00805F319E77EB8533@exs23.ex.nus.edu.sg> hi, friends, Has anybody done something relative to RSVP simulation under ns-2.1b6? thanks & best regards Ge Yu From salehi@isi.edu Fri Jan 7 04:08:14 2000 From: salehi@isi.edu (Nader Salehi) Date: Thu, 6 Jan 2000 20:08:14 -0800 (PST) Subject: new classifier? In-Reply-To: <38746380.89997C8B@ensi.rnu.tn> References: <38746380.89997C8B@ensi.rnu.tn> Message-ID: <200001070408.UAA28253@joy.isi.edu> Did you update tcl/ns-lib.tcl? Nader >>> "Hajer" == Hajer Tounsi writes: Hajer> Hi, I have defined a new classifier (Classifier/IPAgrege), I Hajer> have added in Makefile and Makefile.in the file.o referencing Hajer> the corresponding Tclclass and methods, Hajer> but when I use [new Classifier/IPAgrege] in my simulation, I Hajer> get ..... invalid command Classifier/IPAgrege Hajer> Everything looks to be in order. What am I doing wrong, or what Hajer> should I try that I already haven't? Hajer> Thanks Hajer> Hajer TOUNSI From salehi@isi.edu Fri Jan 7 04:07:19 2000 From: salehi@isi.edu (Nader Salehi) Date: Thu, 6 Jan 2000 20:07:19 -0800 (PST) Subject: address-format In-Reply-To: <200001061539.QAA03552@melle.ffm.fgan.de> References: <200001061539.QAA03552@melle.ffm.fgan.de> Message-ID: <200001070407.UAA28250@joy.isi.edu> The Each node has a 32-bit address space. I'm not sure what you mean though. Nader >>> "Anastasia" == Anastasia Leventi-Peetz writes: Anastasia> Dear ns users, Anastasia> I've been studying the ns Notes and Documentation, after I Anastasia> have installed the ns Packet at my Sun computer. Does Anastasia> anybody know some informations about the maximum value n Anastasia> (number of bits) reserved for the node addresses of the Anastasia> simulator? Is it so that only the value of 32 is allowed Anastasia> in the revised version? Has anyone made experience with Anastasia> something different? I would appreciate some hints Anastasia> A.Leventi-Peetz From miceky_dino@pacific.net.sg Fri Jan 7 14:49:47 2000 From: miceky_dino@pacific.net.sg (Wong Hau Shian) Date: Fri, 07 Jan 2000 22:49:47 +0800 Subject: tkAppInit.O problem (NS) Message-ID: <3.0.6.32.20000107224947.0079a700@pacific.net.sg> Hello, I've got the same problem as Mr Rachad Abdallah. Can you please supply me with the solution to the problem? Thanks. Regards Hau Shian Problem -------------------------------------------------------- Thanks to the quick help i got i was able to overcome my first problem .. not the less .. another problem emerged now .. again runing install gives the following ------------------------------------------------------------- .... /generic/tk.h :77 : X11/Xlib.h No such file or directory make : *** [tkAppInit.o] Error 1 tk8.0.4 makr failed ! Exiting ........... ------------------------------------------------------------- May be I have the wrong tk version or something .. I don't know. Help needed urgently. -------------------------------------- Rashad Hassan Abdallah Siemens AG Information and Communication Networks -------------------------------------- From elkhayat@run.montefiore.ulg.ac.be Fri Jan 7 15:33:51 2000 From: elkhayat@run.montefiore.ulg.ac.be (Ibtissam El Khayat) Date: Fri, 07 Jan 2000 16:33:51 +0100 Subject: mrtproto Message-ID: <3876075F.94BE5639@run.montefiore.ulg.ac.be> Hello I'm new user of ns i 'm trying to simulate a network where only some of nodes reconize the multicast I've tried $ns mrtproto DM $node1 $node2 ... but it does'nt work I used [liste $node.....] does'nt work what should I do?? what's wrong? thanks Ibtissam PS: I've tried to find where mrproto is defined but without succes From haoboy@isi.edu Fri Jan 7 17:32:02 2000 From: haoboy@isi.edu (Haobo Yu) Date: Fri, 7 Jan 2000 09:32:02 -0800 (PST) Subject: tkAppInit.O problem (NS) In-Reply-To: <3.0.6.32.20000107224947.0079a700@pacific.net.sg> Message-ID: Are you using Debian Linux? I recall seeing similar messages in the mailing list before. It sounds like your X is installed in some non-default directories and tk cannot find it. You need to go into tk and do configure using option '--with-x='. See www.scriptics.com for more information on tcl/tk installation. On Fri, 7 Jan 2000, Wong Hau Shian wrote: > Hello, > > I've got the same problem as Mr Rachad Abdallah. Can you please supply me > with the solution to the problem? Thanks. > > Regards > Hau Shian > > Problem > -------------------------------------------------------- > Thanks to the quick help i got i was able to overcome my first problem .. > not the less .. another problem emerged now .. again runing install gives > the following > ------------------------------------------------------------- > .... /generic/tk.h :77 : X11/Xlib.h No such file or directory > make : *** [tkAppInit.o] Error 1 > tk8.0.4 makr failed ! Exiting ........... > ------------------------------------------------------------- > May be I have the wrong tk version or something .. I don't know. Help > needed urgently. > > -------------------------------------- > Rashad Hassan Abdallah > Siemens AG > Information and Communication Networks > -------------------------------------- > > From mallman@grc.nasa.gov Fri Jan 7 21:24:00 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Fri, 07 Jan 2000 16:24:00 -0500 Subject: bug and fix in sack1 Message-ID: <200001072124.QAA06807@lombok-fi.lerc.nasa.gov> I ran across a bug in the sack1 TCP variant this afternoon. When reseting a sack TCP connection the scoreboard is not cleared. In turn, this can really confuse things when sending resumes (or doesn't resume, as in the case I was looking at). As far as I can tell this problem only happens when the connection is in loss recovery when it ends (otherwise the scoreboard is cleared naturally). I am not running the absolute latest version of ns, but I just checked the daily snapshot and it looks like it would still have this problem (but, I didn't compile it and try). The quick change I made that fixed the problem is to add a reset function to the Sack1TcpAgent class as follows: void Sack1TcpAgent::reset () { scb_.ClearScoreBoard(); TcpAgent::reset (); } I didn't check any of the other SACK TCP variants (FACK, rate-halving, etc.), but this could be a problem with them, as well. allman From r7725029@im.ntu.edu.tw Sat Jan 8 05:04:02 2000 From: r7725029@im.ntu.edu.tw (Chen Wen-Jen) Date: Sat, 8 Jan 2000 13:04:02 +0800 (CST) Subject: WRR? Message-ID: hello~~~ guys~~~ I need to use the Weighted Round Robin algorithm in my simulation. Is there any WRR implementation in NS? Jay... From cak@di.ufpe.br Sat Jan 8 22:32:10 2000 From: cak@di.ufpe.br (Carlos Alberto Kamienski) Date: Sat, 8 Jan 2000 20:32:10 -0200 (EDT) Subject: WRR? In-Reply-To: Message-ID: This message is in MIME format. The first part should be readable text, while the remaining parts are likely unreadable without MIME-aware tools. Send mail to mime@docserver.cac.washington.edu for more info. ---559023410-1804928587-947370730=:3008 Content-Type: TEXT/PLAIN; charset=US-ASCII Jain You can implement WRR using CBQ (Class-Based Queue). Configuring CBQ isn't a simple task, and I'm attaching a little example. You can find a more complete example in the test directory. Carlos On Sat, 8 Jan 2000, Chen Wen-Jen wrote: > hello~~~ guys~~~ > > I need to use the Weighted Round Robin algorithm > in my simulation. > > Is there any WRR implementation in NS? > > Jay... > ---559023410-1804928587-947370730=:3008 Content-Type: TEXT/PLAIN; charset=US-ASCII; name="DS-WRR-C.TCL" Content-Transfer-Encoding: BASE64 Content-ID: Content-Description: Content-Disposition: attachment; filename="DS-WRR-C.TCL" DQ0KI0NyZWF0ZSBhIHNpbXVsYXRvciBvYmplY3QNDQpzZXQgbnMgW25ldyBT aW11bGF0b3JdDQ0KDQ0KJG5zIGNvbG9yIDEgR3JlZW4NDQokbnMgY29sb3Ig MiBCbHVlDQ0KJG5zIGNvbG9yIDMgQmxhY2sNDQoNDQoNDQpzZXQgZl9lZiBb b3BlbiBkc19lZi50ciB3XQ0NCnNldCBmX2FmIFtvcGVuIGRzX2FmLnRyIHdd DQ0Kc2V0IGZfYmUgW29wZW4gZHNfYmUudHIgd10NDQoNDQoNDQojRGVmaW5l IGEgJ2ZpbmlzaCcgcHJvY2VkdXJlDQ0KcHJvYyBmaW5pc2gge30gew0NCiAg ICAgICAgZ2xvYmFsIGZfZWYgZl9hZiBmX2JlDQ0KICAgICAgICAjQ2xvc2Ug dGhlIG91dHB1dCBmaWxlcw0NCgljbG9zZSAkZl9lZiANDQoJY2xvc2UgJGZf YWYgDQ0KCWNsb3NlICRmX2JlIA0NCgkNDQoJZXhlYyB4Z3JhcGggZHNfZWYu dHIgZHNfYWYudHIgZHNfYmUudHIgLWdlb21ldHJ5IDcwMHg0MDAgJg0NCiAg ICAgICAgZXhpdCAwDQ0KfQ0NCg0NCnByb2MgcmVjb3JkIHt9IHsNDQogICAg ICAgIGdsb2JhbCBzaW5rX2VmIGZfZWYgc2lua19hZiBmX2FmIHNpbmtfYmUg Zl9iZSANDQoJI0dldCBhbiBpbnN0YW5jZSBvZiB0aGUgc2ltdWxhdG9yDQ0K CXNldCBucyBbU2ltdWxhdG9yIGluc3RhbmNlXQ0NCgkjU2V0IHRoZSB0aW1l IGFmdGVyIHdoaWNoIHRoZSBwcm9jZWR1cmUgc2hvdWxkIGJlIGNhbGxlZCBh Z2Fpbg0NCiAgICAgICAgc2V0IHRpbWUgMC4yNQ0NCiAgICAgICAgI0hvdyBt YW55IGJ5dGVzIGhhdmUgYmVlbiByZWNlaXZlZCBieSB0aGUgdHJhZmZpYyBz aW5rcz8NDQogICAgICAgIHNldCBid19lZiBbJHNpbmtfZWYgc2V0IGJ5dGVz X10NDQogICAgICAgIHNldCBid19hZiBbJHNpbmtfYWYgc2V0IGJ5dGVzX10N DQogICAgICAgIHNldCBid19iZSBbJHNpbmtfYmUgc2V0IGJ5dGVzX10NDQoJ I0dldCB0aGUgY3VycmVudCB0aW1lDQ0KICAgICAgICBzZXQgbm93IFskbnMg bm93XQ0NCgkjQ2FsY3VsYXRlIHRoZSBiYW5kd2lkdGggKGluIE1CaXQvcykg YW5kIHdyaXRlIGl0IHRvIHRoZSBmaWxlcw0NCiAgICAgICAgcHV0cyAkZl9l ZiAiJG5vdyBbZXhwciAkYndfZWYvJHRpbWUqOC8xMDAwMDAwXSINDQogICAg ICAgIHB1dHMgJGZfYWYgIiRub3cgW2V4cHIgJGJ3X2FmLyR0aW1lKjgvMTAw MDAwMF0iDQ0KICAgICAgICBwdXRzICRmX2JlICIkbm93IFtleHByICRid19i ZS8kdGltZSo4LzEwMDAwMDBdIg0NCgkjUmVzZXQgdGhlIGJ5dGVzXyB2YWx1 ZXMgb24gdGhlIHRyYWZmaWMgc2lua3MNDQogICAgICAgICRzaW5rX2VmIHNl dCBieXRlc18gMA0NCiAgICAgICAgJHNpbmtfYWYgc2V0IGJ5dGVzXyAwDQ0K ICAgICAgICAkc2lua19iZSBzZXQgYnl0ZXNfIDANDQoJI1JlLXNjaGVkdWxl IHRoZSBwcm9jZWR1cmUNDQogICAgICAgICRucyBhdCBbZXhwciAkbm93KyR0 aW1lXSAicmVjb3JkIg0NCn0NDQoNDQoNDQoNDQojQ3JlYXRlIGZvdXIgbm9k ZXMNDQpzZXQgbjAgWyRucyBub2RlXQ0NCnNldCBuMSBbJG5zIG5vZGVdDQ0K c2V0IG4yIFskbnMgbm9kZV0NDQpzZXQgbjMgWyRucyBub2RlXQ0NCnNldCBu NCBbJG5zIG5vZGVdDQ0KDQ0KIyBDcmlhIHBlcmZpcyBkZSB0cmFmZWdvDQ0K c2V0IHBfZWYgW25ldyBEU1BlYWtSYXRlUHJvZmlsZV0NDQokcF9lZiBzZXQt c2NvcGUgT05FX1RPX0FOWQ0NCiRwX2VmIHNldCBzcmNfIC0xDQ0KJHBfZWYg c2V0IHBlYWtfIDgwMDAwMA0NCiRwX2VmIHNldCBkZWx0YXRfIDAuMDINDQok cF9lZiBzZXQtRFNDUCBFRg0NCg0NCg0NCiMgQ3JpYSBwZXJmaXMgZGUgdHJh ZmVnbw0NCnNldCBwX2FmIFtuZXcgRFNQZWFrUmF0ZVByb2ZpbGVdDQ0KJHBf YWYgc2V0LXNjb3BlIE9ORV9UT19BTlkNDQokcF9hZiBzZXQgc3JjXyAtMQ0N CiRwX2FmIHNldCBwZWFrXyAzMDAwMDANDQokcF9hZiBzZXQgZGVsdGF0XyAw LjAyDQ0KJHBfYWYgc2V0LURTQ1AgQUYxMQ0NCg0NCiRucyBhdCAwLjAgIiRw X2VmIHN0YXJ0Ig0NCiMkbnMgYXQgMC4wICIkcF9hZiBzdGFydCINDQoNDQoj IGNyaWEgY29uZGljaW9uYWRvcmVzIGRlIHRyYWZlZ28NDQpzZXQgY29uZF9l ZiBbbmV3IERTQ29uZGl0aW9uZXJdDQ0KJGNvbmRfZWYgYWRkLXByb2ZpbGUg JHBfZWYNDQpzZXQgY29uZF9hZiBbbmV3IERTQ29uZGl0aW9uZXJdDQ0KJGNv bmRfYWYgYWRkLXByb2ZpbGUgJHBfYWYNDQoNDQoNDQoNDQojQ3JpYSBvcyBs aW5rcyBlbnRyZSBvcyBub3MNDQokbnMgZHVwbGV4LWxpbmsgJG4wICRuMyAx TWIgMTBtcyBEcm9wVGFpbA0NCiRucyBkdXBsZXgtbGluayAkbjEgJG4zIDFN YiAxMG1zIERyb3BUYWlsDQ0KJG5zIGR1cGxleC1saW5rICRuMiAkbjMgMU1i IDEwbXMgRHJvcFRhaWwNDQokbnMgc2ltcGxleC1saW5rICRuMyAkbjQgMU1i IDEwbXMgQ0JRL1dSUg0NCiRucyBzaW1wbGV4LWxpbmsgJG40ICRuMyAxTWIg MTBtcyBEcm9wVGFpbA0NCnNldCBjYnFsaW5rIFskbnMgbGluayAkbjMgJG40 XQ0NCg0NCg0NCiMgSW5zZXJlIG8gZXNjYWxvbmFkb3IgZGUgZmlsYXMgV1JS L0NCUQ0NCg0NCnNldCB0b3BjbGFzcyBbbmV3IENCUUNsYXNzXQ0NCiMgKHRv cGNsYXNzIGRvZXNuJ3QgaGF2ZSBhIHF1ZXVlKQ0NCiR0b3BjbGFzcyBzZXRw YXJhbXMgbm9uZSAwIDEgYXV0byA4IDIgMA0NCg0NCnNldCBjbGFzc2VfZWYg W25ldyBDQlFDbGFzc10NDQpzZXQgZmlsYV9lZiBbbmV3IFF1ZXVlL0Ryb3BU YWlsXQ0NCiRmaWxhX2VmIHNldCBsaW1pdF8gNDANDQokY2xhc3NlX2VmIGlu c3RhbGwtcXVldWUgJGZpbGFfZWYNDQokY2xhc3NlX2VmIHNldHBhcmFtcyAk dG9wY2xhc3MgdHJ1ZSAwLjQgYXV0byAxIDEgMA0NCg0NCnNldCBjbGFzc2Vf YWYgW25ldyBDQlFDbGFzc10NDQpzZXQgZmlsYV9hZiBbbmV3IFF1ZXVlL0Ry b3BUYWlsXQ0NCiRmaWxhX2FmIHNldCBsaW1pdF8gNjANDQokY2xhc3NlX2Fm IGluc3RhbGwtcXVldWUgJGZpbGFfYWYNDQokY2xhc3NlX2FmIHNldHBhcmFt cyAkdG9wY2xhc3MgdHJ1ZSAwLjQgYXV0byAxIDEgMA0NCg0NCnNldCBjbGFz c2VfYmUgW25ldyBDQlFDbGFzc10NDQpzZXQgZmlsYV9iZSBbbmV3IFF1ZXVl L0Ryb3BUYWlsXQ0NCiRmaWxhX2JlIHNldCBsaW1pdF8gMTUwDQ0KJGNsYXNz ZV9iZSBpbnN0YWxsLXF1ZXVlICRmaWxhX2JlDQ0KJGNsYXNzZV9iZSBzZXRw YXJhbXMgJHRvcGNsYXNzIHRydWUgMC4yIGF1dG8gMSAxIDANDQoNDQoNDQok Y2JxbGluayBpbnNlcnQgJHRvcGNsYXNzDQ0KJGNicWxpbmsgaW5zZXJ0ICRj bGFzc2VfZWYNDQokY2JxbGluayBpbnNlcnQgJGNsYXNzZV9hZg0NCiRjYnFs aW5rIGluc2VydCAkY2xhc3NlX2JlDQ0KDQ0KDQ0KJGNicWxpbmsgYmluZCAk Y2xhc3NlX2VmIDEgOyAjIGZpZCAxDQ0KJGNicWxpbmsgYmluZCAkY2xhc3Nl X2FmIDIgOyAjIGZpZCAyDQ0KJGNicWxpbmsgYmluZCAkY2xhc3NlX2JlIDMg OyAjIGZpZCAzDQ0KDQ0KDQ0KDQ0KDQ0KIyBJbnNlcmUgb3MgY29uZGljaW9u YWRvcmVzIG5hIHJlZGUNDQokbnMgaW5zZXJ0LWNvbmRpdGlvbmVyICRjb25k X2VmICRuMCAkbjMNDQojJG5zIGluc2VydC1jb25kaXRpb25lciAkY29uZF9h ZiAkbjEgJG4zDQ0KDQ0KDQ0KIyBBZ2VudGUgRUYNDQoNDQpzZXQgY2JyX2Vm IFtuZXcgQWdlbnQvQ0JSXQ0NCiRucyBhdHRhY2gtYWdlbnQgJG4wICRjYnJf ZWYNDQokY2JyX2VmIHNldCBwYWNrZXRTaXplXyA1MDANDQokY2JyX2VmIHNl dCBpbnRlcnZhbF8gMC4wMDUNDQokY2JyX2VmIHNldC1EU0NQIEVGDQ0KJGNi cl9lZiBzZXQgZmlkXyAxDQ0KDQ0Kc2V0IHNpbmtfZWYgW25ldyBBZ2VudC9M b3NzTW9uaXRvcl0NDQokbnMgYXR0YWNoLWFnZW50ICRuNCAkc2lua19lZg0N Cg0NCiRucyBjb25uZWN0ICRjYnJfZWYgJHNpbmtfZWYgIA0NCg0NCg0NCiMg Q3JpYSB1bSBhZ2VudGUgVENQL0ZUUCBlIGxpZ2EgY29tIG8gbm9oIG4xIC0g QUYNDQpzZXQgdGNwX2FmIFtuZXcgQWdlbnQvVENQXQ0NCnNldCBzaW5rX2Fm IFtuZXcgQWdlbnQvVENQU2lua10NDQokbnMgYXR0YWNoLWFnZW50ICRuMSAk dGNwX2FmDQ0KJG5zIGF0dGFjaC1hZ2VudCAkbjQgJHNpbmtfYWYNDQokbnMg Y29ubmVjdCAkdGNwX2FmICRzaW5rX2FmDQ0KJHRjcF9hZiBzZXQgZmlkXyAy DQ0KJHRjcF9hZiBzZXQtRFNDUCBBRjExDQ0KDQ0Kc2V0IGZ0cF9hZiBbJHRj cF9hZiBhdHRhY2gtc291cmNlIEZUUF0NDQoNDQojIENyaWEgdW0gYWdlbnRl IFRDUC9GVFAgZSBsaWdhIGNvbSBvIG5vaCBuMiAtIEJFDQ0Kc2V0IHRjcF9i ZSBbbmV3IEFnZW50L1RDUF0NDQpzZXQgc2lua19iZSBbbmV3IEFnZW50L1RD UFNpbmtdDQ0KJG5zIGF0dGFjaC1hZ2VudCAkbjIgJHRjcF9iZQ0NCiRucyBh dHRhY2gtYWdlbnQgJG40ICRzaW5rX2JlDQ0KJG5zIGNvbm5lY3QgJHRjcF9i ZSAkc2lua19iZQ0NCiR0Y3BfYmUgc2V0IGZpZF8gMw0NCiR0Y3BfYmUgc2V0 LURTQ1AgQkUNDQoNDQpzZXQgZnRwX2JlIFskdGNwX2JlIGF0dGFjaC1zb3Vy Y2UgRlRQXQ0NCg0NCg0NCg0NCiNTdGFydCBsb2dnaW5nIHRoZSByZWNlaXZl ZCBiYW5kd2lkdGgNDQokbnMgYXQgMC4wICJyZWNvcmQiDQ0KDQ0KDQ0KJG5z IGF0IDUuMCAiJGNicl9lZiBzdGFydCINDQokbnMgYXQgNS4wICIkZnRwX2Fm IHN0YXJ0Ig0NCiRucyBhdCA1LjAgIiRmdHBfYmUgc3RhcnQiDQ0KDQ0KJG5z IGF0IDU1LjAgIiRjYnJfZWYgc3RvcCINDQokbnMgYXQgNTUuMCAiJGZ0cF9h ZiBzdG9wIg0NCiRucyBhdCA1NS4wICIkZnRwX2JlIHN0b3AiDQ0KDQ0KJG5z IGF0IDYwLjAgImZpbmlzaCINDQoNDQojUnVuIHRoZSBzaW11bGF0aW9uDQ0K JG5zIHJ1bg0NCg== ---559023410-1804928587-947370730=:3008-- From floyd@aciri.org Sun Jan 9 04:11:11 2000 From: floyd@aciri.org (Sally Floyd) Date: Sat, 08 Jan 2000 20:11:11 -0800 Subject: bug and fix in sack1 Message-ID: <200001090411.UAA13438@elk.aciri.org> Mark - >I ran across a bug in the sack1 TCP variant this afternoon. When >reseting a sack TCP connection the scoreboard is not cleared. Thanks for the bug report and the fix - it will be in the nightly snapshot. >I didn't check any of the other SACK TCP variants (FACK, >rate-halving, etc.), but this could be a problem with them, as well. I will add it to the fack code also. The rate-halving code hasn't been incorporated into the NS distribution yet. - Sally -------------------------------- http://www.aciri.org/floyd/ -------------------------------- From mfrumin@nepal.stanford.edu Sun Jan 9 06:31:16 2000 From: mfrumin@nepal.stanford.edu (Mike Frumin) Date: Sat, 8 Jan 2000 22:31:16 -0800 (PST) Subject: Cant make make-lan work Message-ID: Hi. I've seen this problem posted before, but never the solution. when using the standard call to make-lan as provided in the ns users manual (i.e. $ns make-lan "$node1 $node 2" 100 Mb 1ms LL Queue/DropTail Mac/Csma/Cd) I get ugly errors. It gives an error with the final Mac/Csma/Cd command. silly. and if I take out the list of nodes (i.e. put in the empty string) I get other nonsense errors. any thought? I'm using 2.1b5 thanks a lot, Mike From guru0109@palgong.knu.ac.kr Sun Jan 9 13:09:23 2000 From: guru0109@palgong.knu.ac.kr (=?euc-kr?B?w9a/tbz2?=) Date: Sun, 9 Jan 2000 22:09:23 +0900 Subject: Fair Queueing Scheduling Algorithms available??? Message-ID: <00a001bf5aa2$bfe01a00$d50ee69b@knu.ac.kr> Hi all. I search NS mailing archive these question and contact sudhir but haven't got no any reaply... : ( So I ask again... Is there anyone who implemeted some scheduling algorithms for NS2??? Especially I need SCFQ or STFQ(Start Time Fair Queueing). And if anyone did, can you share some codes to the public domain? I'm new user at NS so it's so hard to put my effort on implementing these scheduling algorithms. Any suggestions or advices are welcome very much! Thanks A Lot! ______________________________________________________ ChoiYoungSoo e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 TeNet Lab : http://tenet.knu.ac.kr/ From huang@tik.ee.ethz.ch Mon Jan 10 10:18:10 2000 From: huang@tik.ee.ethz.ch (Polly Huang) Date: Mon, 10 Jan 2000 11:18:10 +0100 (CET) Subject: Multicast Questions In-Reply-To: <08092713182501.00880@jamester.snu.ac.kr> Message-ID: On Sat, 27 Sep 2008, jamester wrote: > Hello. > > My first question : > > It seems like "$ns set-address-format expanded" and multicast > do not go together. When I use expanded address I get errors similar to "_o57: > no target for slot 128". This error occurs when I generate only 100 nodes with > expanded address format, but not without expanded address format. Anyone knows > anything about this? > > My second question : > Is there any way I can see routes from source to destination nodes in > multicast? I tried following : > > set mproto DM > set mrthandle [$ns mrtproto $mproto {}] > ........ > $ns join-group receiver_agent group > $ns get-mcast-tree $node($src) $group #src is srouce node id > > I do not get multicast tree but following error messages. > > ns: _o3 get-mcast-tree _o53 0x8002]: can't read "GROUP_": no such variable ^^^^^^ James, 0x8002 is considerred a unicast address when used with 'expanded' addressing. It is recommended to use 'Node allocaddr' instead of manually configuring mcast addresses (unless you know the address format very well). Cheers, -Polly > while executing > "error "No replicator for $GROUP_ at [$src id]"" > (procedure "_o3" line 7) > (Simulator get-mcast-tree line 7) > invoked from within > "_o3 get-mcast-tree _o53 0x8002]" > > Any help is greatly appreciated. > Thanks in advance. > > -James Kwon > > > > From L.Wood@eim.surrey.ac.uk Mon Jan 10 11:34:57 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 10 Jan 2000 11:34:57 +0000 (GMT) Subject: Multicast Questions In-Reply-To: <08092713182501.00880@jamester.snu.ac.kr> Message-ID: On Sat, 27 Sep 2008, jamester wrote: > My first question : > > It seems like "$ns set-address-format expanded" and multicast > do not go together. > When I use expanded address I get errors similar to "_o57: > no target for slot 128". This error occurs when I generate > only 100 nodes with expanded address format, but not without > expanded address format. Anyone knows anything about this? What version of ns are you using? I spent some time discovering and pointing this particular gotcha out on the mailing list circa 2.1b3[*] Instead of set-address-format, stick to Node expandaddr. Take a look through the ns mailing list archive[+] http://www-mash.CS.berkeley.EDU/dist/archive/ns-users/9809/0066.html http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9901/0072.html There's a forwarded note giving some indication why (no idea where that came from) at: http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9901/0084.html and addressing has been improved more recently (although afaik the various addressing modes are not all some aspect of hierarchical addressing - one level of hierarchy would give a flat space, and one overall type of addressing might simplify interactions), so this should be less of a problem in recent ns. When in doubt, try Node expandaddr. For your second problem below, make sure you're using Node allocaddr if you're using expanded addressing - 0x8002 is only valid as a multicast address for the slowly-being-deprecated small address space when you have 128 nodes or less. Assuming you aren't using expanded addressing... ...well, I've never seen get-mcast-tree before - where is that defined? I can't find it in my recent snapshot or 2.1b5 or the docs. thanks, L. [*] these days I tend to stick to pointing out that DV routing is broken, that the only person who has got it working again can't remember how he did it, and can we have it functional again for 2.1b6, please? [+] hang on, ht://dig currently isn't generating header/footer parts of search result pages when it's not returning no data, so you're limited to the first ten results; you'd never be able to turn this stuff up. PGP > My second question : > Is there any way I can see routes from source to destination nodes in > multicast? I tried following : > > set mproto DM > set mrthandle [$ns mrtproto $mproto {}] > ........ > $ns join-group receiver_agent group > $ns get-mcast-tree $node($src) $group #src is srouce node id > > I do not get multicast tree but following error messages. > > ns: _o3 get-mcast-tree _o53 0x8002]: can't read "GROUP_": no such variable > while executing > "error "No replicator for $GROUP_ at [$src id]"" > (procedure "_o3" line 7) > (Simulator get-mcast-tree line 7) > invoked from within > "_o3 get-mcast-tree _o53 0x8002]" > > Any help is greatly appreciated. > Thanks in advance. > > -James Kwon From jamester@dacl3.snu.ac.kr Sun Jan 9 11:50:44 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Sun, 9 Jan 2000 20:50:44 +0900 Subject: Multicast Questions References: Message-ID: <00010920535500.01085@jamester.snu.ac.kr> I believe get-mcast-tree is defined in ns/ns-xxx/tcl/mcast/ns-mcast.tcl. It is defined as following. Simulator instproc get-mcast-tree { src grp } { $self instvar link_ treeLinks_ # iif == -2: from local set tmp [$src getRepByIIF $src $grp -2] if {$tmp == ""} { $self flush-trace error "No replicator for $grp at [$src id]" } set sid [$src id] lappend repList $tmp lappend nodeList $src while {[llength $repList] > 0} { set h [lindex $repList 0] set parent [lindex $nodeList 0] set pid [$parent id] set closeNodes($pid) 1 set repList [lreplace $repList 0 0] set nodeList [lreplace $nodeList 0 0] set slots [$h slots] # puts "$slots" foreach sl $slots { set child [$self find-next-child $parent $sl] if {$child == ""} { puts "No children found for ($parent, $sl)" continue } set cid [$child id] if [info exists closeNodes($cid)] { error "Loop: node $cid already in the tree" } # shouldn't go upstream if ![info exists link_($pid:$cid)] { error "Found non-existent link ($pid:$cid)"; } lappend nodeList $child puts "Link ($pid:$cid) found." lappend treeLinks $pid:$cid $link_($pid:$cid) set treeLinks_($pid:$cid) $link_($pid:$cid) set iif [[$link_($pid:$cid) set ifaceout_] id] if {$iif == -1} { puts "iif == -1" } lappend repList [$child getRepByIIF $src $grp $iif] } } return $treeLinks } It seems as if all the paticipants of multicast should join group before this proc is called. I just can not get this proc to work correctly. By the way, I am using 2.1b4 for some special reasons. Thanks -James Kwon Mon, 10 Jan 2000俊, Lloyd Wood 累己茄 臂: > On Sat, 27 Sep 2008, jamester wrote: > > > My first question : > > > > It seems like "$ns set-address-format expanded" and multicast > > do not go together. > > > When I use expanded address I get errors similar to "_o57: > > no target for slot 128". This error occurs when I generate > > only 100 nodes with expanded address format, but not without > > expanded address format. Anyone knows anything about this? > > What version of ns are you using? I spent some time > discovering and pointing this particular gotcha out on the mailing > list circa 2.1b3[*] > > Instead of set-address-format, stick to Node expandaddr. > > Take a look through the ns mailing list archive[+] > > http://www-mash.CS.berkeley.EDU/dist/archive/ns-users/9809/0066.html > http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9901/0072.html > > There's a forwarded note giving some indication why (no idea where > that came from) at: > http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9901/0084.html > > and addressing has been improved more recently (although afaik the > various addressing modes are not all some aspect of hierarchical > addressing - one level of hierarchy would give a flat space, and one > overall type of addressing might simplify interactions), so this > should be less of a problem in recent ns. > > When in doubt, try Node expandaddr. > > For your second problem below, make sure you're using Node allocaddr > if you're using expanded addressing - 0x8002 is only valid as a > multicast address for the slowly-being-deprecated small address space > when you have 128 nodes or less. Assuming you aren't using expanded > addressing... > > ...well, I've never seen get-mcast-tree before - where is that > ddefined? I can't find it in my recent snapshot or 2.1b5 or the docs. > > thanks, > > L. > > [*] these days I tend to stick to pointing out that DV routing is > broken, that the only person who has got it working again can't > remember how he did it, and can we have it functional again for > 2.1b6, please? > > [+] hang on, ht://dig currently isn't generating header/footer > parts of search result pages when it's not returning no data, so > you're limited to the first ten results; you'd never be able to > turn this stuff up. > > PGP > > > My second question : > > Is there any way I can see routes from source to destination nodes in > > multicast? I tried following : > > > > set mproto DM > > set mrthandle [$ns mrtproto $mproto {}] > > ........ > > $ns join-group receiver_agent group > > $ns get-mcast-tree $node($src) $group #src is srouce node id > > > > I do not get multicast tree but following error messages. > > > > ns: _o3 get-mcast-tree _o53 0x8002]: can't read "GROUP_": no such variable > > while executing > > "error "No replicator for $GROUP_ at [$src id]"" > > (procedure "_o3" line 7) > > (Simulator get-mcast-tree line 7) > > invoked from within > > "_o3 get-mcast-tree _o53 0x8002]" > > > > Any help is greatly appreciated. > > Thanks in advance. > > > > -James Kwon From L.Wood@eim.surrey.ac.uk Mon Jan 10 12:05:03 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 10 Jan 2000 12:05:03 +0000 (GMT) Subject: Multicast Questions In-Reply-To: <00010920535500.01085@jamester.snu.ac.kr> Message-ID: On Sun, 9 Jan 2000, jamester wrote: > I believe get-mcast-tree is defined in ns/ns-xxx/tcl/mcast/ns-mcast.tcl. not any more. > It seems as if all the paticipants of multicast should join group before this > proc is called. I just can not get this proc to work correctly. > By the way, I am using 2.1b4 for some special reasons. http://www-mash.cs.berkeley.edu/cgi-bin/cvsweb/ns-2/tcl/mcast/ns-mcast.tcl that got commented out (1.34) and then removed (1.36). Node expandaddr, and good luck. L. PGP From Tarik Alj Mon Jan 10 13:55:56 2000 From: Tarik Alj (Tarik Alj) Date: Mon, 10 Jan 2000 08:55:56 -0500 (EST) Subject: Cant make make-lan work Message-ID: <200001101355.IAA28587@opale.INRS-Telecom.UQuebec.CA> Please use mac-802_3 instead. MAC/CSMA/CD is not valid anymore. Also consult the cvs tree to see if your vlan.tcl file is up to date. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Sat, 8 Jan 2000 22:31:16 -0800 (PST) > From: Mike Frumin > To: ns-users@mash.cs.berkeley.edu > Subject: Cant make make-lan work > MIME-Version: 1.0 > > Hi. I've seen this problem posted before, but never the solution. > when using the standard call to make-lan as provided in the ns users > manual (i.e. $ns make-lan "$node1 $node 2" 100 Mb 1ms LL Queue/DropTail > Mac/Csma/Cd) I get ugly errors. It gives an error with the final > Mac/Csma/Cd command. silly. and if I take out the list of nodes (i.e. put > in the empty string) I get other nonsense errors. > any thought? I'm using 2.1b5 > > thanks a lot, > Mike Tarik From nieri@magna.ing.unipi.it Mon Jan 10 14:18:44 2000 From: nieri@magna.ing.unipi.it (Fabio Nieri) Date: Mon, 10 Jan 2000 15:18:44 +0100 (CET) Subject: Wireless Error Model Message-ID: Hi, I'm sorry for I've already asked this question a month ago but seeing that it still remains unaswered I post it again. I'm trying to simulate a wireless network with FTP traffic running on it and would like to insert a two state Markovian error model at the physical layer level. I cannot find anything useful in the archives nor in the documentation; all seems to be related to link errors. Can anyone suggest how to insert such an error? Thanks in advance Fabio Nieri From barenco@dit.upm.es Mon Jan 10 14:28:53 2000 From: barenco@dit.upm.es (Claudia Jacy Barenco) Date: Mon, 10 Jan 2000 15:28:53 +0100 Subject: Variuous links on the same interface Message-ID: <001b01bf5b77$07967540$43d7523e@fofinha> Dear users, Im trying to configure three simplex links between the same nodes but I receive an error. Im wondering if is not possible to establish various logical links on the same physical interface. Thanks for attention. *************************************** Cl醬dia Jacy Barenco Polythecnic University of Madrid Telematic Systems E.T.S.I. Telecommunication Madrid - Spain *************************************** From Hajer.Tounsi@ensi.rnu.tn Mon Jan 10 15:17:19 2000 From: Hajer.Tounsi@ensi.rnu.tn (Hajer Tounsi) Date: Mon, 10 Jan 2000 16:17:19 +0100 Subject: new classifier? References: <38746380.89997C8B@ensi.rnu.tn> <000b01bf588d$82229240$f7af15a5@ericwong> Message-ID: <3879F7FF.C936F086@ensi.rnu.tn> Hi, > I have a question. are you going to use 2 or more different types of > classifier in your topology? the reason why I ask is I need 2 >different types of nodes in my network based on my 2 classifier-A and >B. how do I do that?? yes, I am going to use 2 types of nodes in my topology. a simple node with the Classifier/Addr and a new type of node which i have create (NodeA) and which will use another classifier-A which I also define, but the problem is when I use in the NodeA instproc mk-default-classifier [new Classifier/NodeA] when I use create_nodeA , I have an error : Invalid command name Classifier/NodeA. some help? thanks hajer From L.Wood@eim.surrey.ac.uk Mon Jan 10 21:13:02 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 10 Jan 2000 21:13:02 +0000 (GMT) Subject: bug report - TCPSink/*, packetSize_, headers and recvBytes() Message-ID: I ran into an anomaly in trying to get an end-application receiver view of a tcp transfer for an accurate application-level view of transfer time/throughput. It seems to affect all one-way TCP variants in ns up to and including the current snapshot, but doesn't affect FullTcp, which does the Right Thing. One-way TCPSinks consider packetSize_ (confusingly defined as size_ rather than packetSize_ in C++, alas) to be the segment size as well, without taking header overhead into account at any point after receiving a packet and before passing received segment data from the window up to recvBytes() for an application to take. I was expecting a difference of multiples of 40 or so with respect to n*(packetSize_) in numtoDeliver to account for the necessary TCP and IP header overhead. You can see this by adding to tcp-sink.{cc,h}: void TcpSink::recvBytes(int bytes) { // Would it be possible to have #define NOW available // across all of ns globally? I'm really quite taken with it, // and surprised when it's not available without another // #include. double now = Scheduler::instance().clock(); // show that chunks being passed up to application are // n*foo from set packetSize_ foo - it's now segment size! printf("%f\t%d\n",now,bytes); } do set packetSize_ 1000 before initiating an FTP agent over a TCP connection, and you'll see chunks of 1000 bytes being handed up to the application. So, packetSize_ has become size_ which has mutated to segment size as far as the sink receiver and the application it would be handing data to are concerned, and the values passed to recvBytes() via numBytes and numToDeliver are somewhat misleading. (If there are better ways of recording transfer rates at the receiver, I'm all ears.) Since as far as the tracefiles and ns are concerned, packetSize_ really *is* packet size (as nsDoc 25.1 states), numBytes and numtoDeliver in tcp-sink.cc are at least multiples of 40 too large for recvBytes() (TCP options are probably moot since it's one-way), and getting an accurate measure of the rate of data being delivered to the application via recvBytes() means that postprocessing adjustment for header overhead would be needed to correct this for one-way TCP - once you'd noticed it. Ouch. I think you could use a variable similar to the bound tcpip_base_hdr_size_ variable in tcp.h - which seems to be used to adjust for headers on the way out in the one-way sender - to handle this by subtraction; better might be to use hlen_ in hdr_tcp for one-way TCPs as well. (btw, if you're setting packetSize_ to less than 40, things get extremely odd. An impossible 30-byte tcp packet will actually send 30 bytes up to recvBytes() for the application - imo this needs guarding/warning against, and a simple tcpip_base_hdr_size_ equivalent subtraction to give the correct value would go negative and take bytes _back_ from the receiving application...) Adding a similar recvBytes() method to tcp-full.cc shows that FullTcp is doing the Right Thing as far as header overhead is concerned - multiples of segsize_ are passed up, and the tracefile records a value 40 bytes bigger, including the header overhead just as you'd expect. Anyone comparing throughputs from FullTcp/reality and one-way TCPs for the same applications using the apparently-same recvBytes() for an identical API could be tripped up by this (and I was). 40 bytes/packet doesn't seem like much, but over long flows it adds up, and if your application is depending on chunks of data to arrive before doing any processing... [How much performance overhead would turning the bytes received into a bound variable - to be read off at intervals - add?] cheers, L. I'm blaming this all on calling it size_, you know. PGP From kpsounis@leland.Stanford.EDU Tue Jan 11 00:27:03 2000 From: kpsounis@leland.Stanford.EDU (Konstantinos Psounis) Date: Mon, 10 Jan 2000 16:27:03 -0800 Subject: ns-allinone-2.1b5 installation problem on solaris 7 Message-ID: <387A78D7.BE12C6D2@leland.stanford.edu> Hi all, I have the 1st installation problem reported in the bug fixes page with the ns-allinone-2.1b5, while truing to install it under solaris 7: "Problem: When installing ns-allinone or ns, I get an error message that includes configure: error: can not run test program while cross compiling" I relaize this is an error in tcl (probably in my gcc) but since I installed the gcc-2.95.1 from the sunfreeware.com site (package), I guess that other people had exactly the same problem with me and I wanted to ask what they did to fix it: 1)run gcc in non cross compile mode. In this case ns is installed but nam doesn't work (at least in my system). 2)install another gcc version (maybe the 2.95.2)? 3)use another ns-allinone version (maybe the 2.1b4)? Thanks a lot, Konstantinos Psounis From salehi@isi.edu Tue Jan 11 00:37:47 2000 From: salehi@isi.edu (Nader Salehi) Date: Mon, 10 Jan 2000 16:37:47 -0800 (PST) Subject: Variuous links on the same interface In-Reply-To: <001b01bf5b77$07967540$43d7523e@fofinha> References: <001b01bf5b77$07967540$43d7523e@fofinha> Message-ID: <200001110037.QAA18222@joy.isi.edu> You should not have a problem. If this is a part of a bigger program, then simplify the program to contain just the node creation and establishing links between them. Nader >>> "Claudia" == Claudia Jacy Barenco writes: Claudia> Dear users, Im trying to configure three simplex links Claudia> between the same nodes but I receive an error. Im wondering Claudia> if is not possible to establish various logical links on the Claudia> same physical interface. Thanks for attention. Claudia> *************************************** Cl醬dia Jacy Barenco Claudia> Polythecnic University of Madrid Telematic Systems Claudia> E.T.S.I. Telecommunication Madrid - Spain Claudia> *************************************** From jamester@dacl3.snu.ac.kr Tue Jan 11 02:28:40 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Tue, 11 Jan 2000 11:28:40 +0900 Subject: Multicast Questions References: Message-ID: <00011111325700.01353@jamester.snu.ac.kr> > James, > > 0x8002 is considerred a unicast address when used with 'expanded' > addressing. It is recommended to use 'Node allocaddr' instead of manually > configuring mcast addresses (unless you know the address format very > well). > > Cheers, > -Polly > > > while executing > > "error "No replicator for $GROUP_ at [$src id]"" > > (procedure "_o3" line 7) > > (Simulator get-mcast-tree line 7) > > invoked from within > > "_o3 get-mcast-tree _o53 0x8002]" > > Dear Polly Huang Using Node allocaddr eliminated error of my first question - no target for slot error. However, it did not solve problem with getting multicast tree. I am using "ns 2.1b4". I notice that CtrMcastComp.tcl has routine for computing multicast tree. Is there any way to find out which nodes are in the tree? Thanks -James Kwon From johnh@isi.edu Tue Jan 11 05:07:42 2000 From: johnh@isi.edu (John Heidemann) Date: Mon, 10 Jan 2000 21:07:42 -0800 Subject: Question In-Reply-To: Message from Arun Gandhi of "Thu, 06 Jan 2000 19:21:40 EST." <38753194.A77BC1E@cisunix.unh.edu> Message-ID: <200001110507.VAA03939@dash.isi.edu> >In the example I have I need to generate nodes after certain time that is after running simulation for some time.How can I > >dynamically generate a node.(source,intermediate or destination). > >I am trying to change topology dynamically.(at run time and see how is my network behaving) There are examples that do this in the distribution. See many_tcp.tcl, for examples. -John Heidemann From guru0109@palgong.knu.ac.kr Tue Jan 11 05:26:24 2000 From: guru0109@palgong.knu.ac.kr (guru0109) Date: Tue, 11 Jan 2000 14:26:24 +0900 Subject: Creat MultiSession In One Node Message-ID: <387ABEFF.E0C22F60@palgong.knu.ac.kr> Hi. I modify test-drr.tcl script for my simulation. I want to use same topology. But want to create many sessions in one nodes (maybe about 500 sessions). so i decide to use for loop in tcl script and modify drr example script. but occur these erros --- ns: _o89 start: (_o89 cmd line 1) invoked from within "_o89 cmd start" invoked from within "catch "$self cmd $args" ret" (procedure "_o89" line 2) (SplitObject unknown line 2) invoked from within "_o89 start" --- Here Is My test script --- snip --- 8 # Run this script as 9 # ../../ns test-drr.tcl 10 # or 11 # ../../ns test-drr.tcl setmask 12 # After running this script cat "out" to view the flow stats 13 14 #default Mask parameter 15 set MASK 0 --- snip --- 24 # Create a simple four node topology: 25 # 26 # n2 27 # \ 28 # 10Mb,0.01ms \ 1Mb,0.01ms 29 # n0 --------- n1 30 # 10Mb,0.01ms / 31 # / 32 # n3 33 34 set n0 [$ns node] 35 set n1 [$ns node] 36 set n2 [$ns node] 37 set n3 [$ns node] 38 39 set nf [open out.nam w] 40 set f [open out.tr w] 41 42 # $ns trace-all $f 43 $ns namtrace-all $nf 44 45 $ns duplex-link $n0 $n1 1.0Mb 0.01ms DRR 46 $ns duplex-link $n2 $n0 10.0Mb 0.01ms DropTail 47 $ns duplex-link $n3 $n0 10.0Mb 0.01ms DropTail 48 49 # trace the bottleneck queue 50 $ns trace-queue $n0 $n1 $f 51 52 Simulator instproc get-link { node1 node2 } { 53 $self instvar link_ 54 set id1 [$node1 id] 55 set id2 [$node2 id] 56 return $link_($id1:$id2) 57 } 58 59 #Alternate way for setting parameters for the DRR queue 60 set l [$ns get-link $n0 $n1] 61 set q [$l queue] 62 63 $q mask $MASK 64 $q blimit 25000 65 $q quantum 500 66 $q buckets 7 67 68 # create 6 cbr sources : 69 70 ################################ Modification Start 73 for {set i 0} {$i < 5} {incr i} { 74 set udp($i) [new Agent/UDP] 75 } 76 77 for {set i 0} {$i < 3} {incr i} { 78 $ns attach-agent $n2 $udp($i) 79 } 80 for {set i 3} {$i < 5} {incr i} { 81 $ns attach-agent $n3 $udp($i) 82 } 83 84 for {set i 0} {$i < 5} {incr i} { 85 set cbr($i) [new Agent/UDP] 86 } 87 88 for {set i 5} {$i < 5} {incr i} { 89 $cbr($i) attach-agent $udp($i) 90 $cbr($i) set packet_size_ 250 91 $cbr($i) set interval_ 8ms 92 $cbr($i) set random_ 1 93 } 94 95 set lm0 [new Agent/Null] 96 $ns attach-agent $n1 $lm0 97 98 for {set i 0} {$i < 5} {incr i} { 99 $ns connect $udp($i) $lm0 100 } 101 for {set i 0} {$i < 5} {incr i} { 102 $ns at 0.0 "$cbr($i) start" 103 } 104 for {set i 0} {$i < 5} {incr i} { 105 $ns at 1.0 "$cbr($i) stop" 106 } 107 108 $ns at 2.0 "close $f;finish" 109 110 proc finish {} { --- snip --- Thx in advance! From huang@tik.ee.ethz.ch Tue Jan 11 09:44:14 2000 From: huang@tik.ee.ethz.ch (Polly Huang) Date: Tue, 11 Jan 2000 10:44:14 +0100 (CET) Subject: Multicast Questions In-Reply-To: <00011111325700.01353@jamester.snu.ac.kr> Message-ID: James, Simulator instproc get-mcast-tree { src grp } { $self instvar link_ treeLinks_ # iif == -2: from local set tmp [$src getRepByIIF $src $grp -2] if {$tmp == ""} { $self flush-trace error "No replicator for $grp at [$src id]" ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ } It looks like the second error you encountered is due to no 'group replicator for the (src grp) given'. You should make sure you actually have receivers joined to the 'grp' and the data sent from 'src'. According to the script you attached, at least 'group' should be '$group' (if that's not a typo in only the email) and you should probably call 'get-mcast-tree' later in the simulation to give the DM and data source some time to set up the replicators (usually the (S,G) replicators are not created in the setup phase). Just a couple speculation what the problems might be. Hope this helps. In general, it's good to create a simple script, visualize it with nam, (just to make sure it works and you're not overlooking something) and then try to add more functions for your purpose. set mproto DM set mrthandle [$ns mrtproto $mproto {}] ........ $ns join-group receiver_agent group $ns get-mcast-tree $node($src) $group #src is srouce node id Cheers, -Polly On Tue, 11 Jan 2000, jamester wrote: > > James, > > > > 0x8002 is considerred a unicast address when used with 'expanded' > > addressing. It is recommended to use 'Node allocaddr' instead of manually > > configuring mcast addresses (unless you know the address format very > > well). > > > > Cheers, > > -Polly > > > > > while executing > > > "error "No replicator for $GROUP_ at [$src id]"" > > > (procedure "_o3" line 7) > > > (Simulator get-mcast-tree line 7) > > > invoked from within > > > "_o3 get-mcast-tree _o53 0x8002]" > > > > > Dear Polly Huang > > Using Node allocaddr eliminated error of my first question - no target for slot > error. However, it did not solve problem with getting multicast tree. > I am using "ns 2.1b4". I notice that CtrMcastComp.tcl has routine for > computing multicast tree. Is there any way to find out which nodes are in the > tree? > > Thanks > > -James Kwon > From jamester@dacl3.snu.ac.kr Tue Jan 11 10:28:41 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Tue, 11 Jan 2000 19:28:41 +0900 Subject: Multicast tree Message-ID: <00011119290701.02188@jamester.snu.ac.kr> Hi. I am trying to find out which nodes are part of multicast tree. I create 1000 nodes and randomly pick one source and 10 ~ 100 destination nodes. I need to find out which nodes are intermediate nodes in the multicst tree from source to destinations so that I can attach special agents to those nodes. Any ideas? Thanks in advance -James Kwon From engp9264@nus.edu.sg Tue Jan 11 13:25:54 2000 From: engp9264@nus.edu.sg (engp9264@nus.edu.sg) Date: Tue, 11 Jan 2000 21:25:54 +0800 Subject: TracedDouble Message-ID: <387B2F62.EF0BDDD4@nus.edu.sg> When I define a variable as TracedDouble, there is a segmentation error. What is wrong? I use ns-2.1b3, tclcl-1.0b6. I also use tclcl-1.0b8. There is no use. Thank your help. From rosblom@yahoo.com Tue Jan 11 15:55:44 2000 From: rosblom@yahoo.com (=?iso-8859-1?q?Andreas=20Rosblom?=) Date: Tue, 11 Jan 2000 07:55:44 -0800 (PST) Subject: RSVP/ns simulation error!! Message-ID: <20000111155544.10595.qmail@web1203.mail.yahoo.com> I'm running RSVP/ns (updated by Sean Murphy) with ns-2b5 on Redhat Linux 5.2. I successfully simulated the file "rsvp_conf.tcl", but when I try the three other files included in the RSVP/ns package (rsvp_distinct.tcl, rsvp_large.tcl, rsvp_merge.tcl), I get a segmentation fault (core dumped). Why? Thanks for any help. /Andreas __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com From agandhi@cisunix.unh.edu Tue Jan 11 21:44:17 2000 From: agandhi@cisunix.unh.edu (Arun Gandhi) Date: Tue, 11 Jan 2000 16:44:17 -0500 Subject: Ques Message-ID: <387BA431.F5DFB2C@cisunix.unh.edu> I don't understand this expression in one of the example calculating queue-delay. set delay [expr 8 * [$bytesinteg set sum_] / [[$link link] set bandwidth_]] Regards, Arun -- Arun Gandhi M.S. (Computer Science) Phone : 603-862-7614 (Res) University of New Hampshire, 603-862-3786 (Off) Durham, NH 03824 Email : agandhi@cisunix.unh.edu From alhussein_abouzeid@hotmail.com Tue Jan 11 22:26:34 2000 From: alhussein_abouzeid@hotmail.com (Alhussein Abouzeid) Date: Tue, 11 Jan 2000 22:26:34 GMT Subject: Wireless Error Model Message-ID: <20000111222634.67392.qmail@hotmail.com> Check out the contributed modules page. Hussein. >From: Fabio Nieri >To: ns-users@mash.cs.berkeley.edu >Subject: Wireless Error Model >Date: Mon, 10 Jan 2000 15:18:44 +0100 (CET) > >Hi, >I'm sorry for I've already asked this question a month ago but seeing that >it still remains unaswered I post it again. I'm trying to simulate a >wireless network with FTP traffic running on it and would like to insert a >two state Markovian error model at the physical layer level. >I cannot find anything useful in the archives nor in the documentation; >all seems to be related to link errors. >Can anyone suggest how to insert such an error? > >Thanks in advance > >Fabio Nieri > > ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From zl@imail.com.cn Wed Jan 12 01:21:54 2000 From: zl@imail.com.cn (zl@imail.com.cn) Date: Wed, 12 Jan 2000 09:21:54 +0800 (CST) Subject: Agent/RTP or Agent/UDP Message-ID: Dear Sir: I am using ns-2 to simulate a simple RTP flow. But I am not sure about the agent. Application is Exponential,it has a parameter called packet_size_; Agent is Agent/RTP,it has a parameter called Packetsize_. My application packet is 160-byte long. Whether I set Packetsize_ at Agent/RTP or not? If it should be set,set it as 160 or 160+12=172? I search the tracing result for packetsize and find the packetsize_ in Agent/RTP may be decisive. Or, I must use Agent/UDP instead of Agent/RTP,but how to add RTP overhead? Thanks for reading my question.Better give me some suggestion. Zhang Lei ---------------------------------------------- 欢迎您使用 百家商务电子邮件系统 http://www.email.com.cn Welcome to E-mail business system From salehi@isi.edu Wed Jan 12 01:53:39 2000 From: salehi@isi.edu (Nader Salehi) Date: Tue, 11 Jan 2000 17:53:39 -0800 (PST) Subject: Multicast tree In-Reply-To: <00011119290701.02188@jamester.snu.ac.kr> References: <00011119290701.02188@jamester.snu.ac.kr> Message-ID: <200001120153.RAA00447@joy.isi.edu> Look at the examples. There are some scripts which make use of labels and colors. Nader >>> "jamester" == jamester writes: jamester> Hi. I am trying to find out which nodes are part of jamester> multicast tree. I create 1000 nodes and randomly pick one jamester> source and 10 ~ 100 destination nodes. I need to find out jamester> which nodes are intermediate nodes in the multicst tree from jamester> source to destinations so that I can attach special agents jamester> to those nodes. Any ideas? jamester> Thanks in advance jamester> -James Kwon From yanyu@catarina.usc.edu Wed Jan 12 05:12:45 2000 From: yanyu@catarina.usc.edu (Yan Yu) Date: Tue, 11 Jan 2000 21:12:45 -0800 (PST) Subject: header size in ns Message-ID: Hello, all I am doing simulation using DSR in ns-2. I am trying to add some extra info in the source route header. so I also modify th hdr_sr::size() to count the extra filed in sr_addr. BUt the simulation behaves wierd after I made those change. If I only change( add a field to) the sr_addr which is included in hdr_sr, but did not change hdr_size(), which I think it is not right, but the simulation seems working ok. So I am wondering what change I need to make when I add some field in packet header, i.e. hdr_sr. Do I need to change hdr_sr::size()? Is it to count the size of the route header? Your any comments would be very helpful and appreciated. Thanks in advance! yan From nkjiang@hotmail.com Wed Jan 12 20:30:54 2000 From: nkjiang@hotmail.com (NingKang Jiang) Date: Wed, 12 Jan 2000 14:30:54 CST Subject: ASk for help Message-ID: <20000112063054.74172.qmail@hotmail.com> Hi, How do you do! I want to develop a MPLS module using ns-2 to do some research work about MPLS technology. Do you know someone else is doing the same thing? If so,I want to discuss with him(her). Thanks, My email: nkjiang@hotmail.com Ningkang Jiang 2000.1.12 ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From nsuser@nepal.stanford.edu Wed Jan 12 09:56:14 2000 From: nsuser@nepal.stanford.edu (nsuser@nepal.stanford.edu) Date: Wed, 12 Jan 2000 01:56:14 -0800 (PST) Subject: LossMonitor and multicast Message-ID: I'm trying to use an Agent/LossMonitor subclass to monitor losses at a node that is subscribed to any number of multicast groups. The problem is, it reports losses (by invoking the log-loss method) when the total bandwidth of all subscribed groups is well below the link capacity. (There is no other network traffic) My only thought is the presence of the multiple groups (and hence group addresses) is confusing the LossMonitor (it might be comparing sequence numbers from different streams) into thinking there are losses when there are not. Does anyone know of a way to have a lossMonitor that is intelligent about multiple streams? I would think Agent/LossMonitor would do this, because it doesn't seem to useful otherwise, but I must be missing something. Any help is greatly appreciated. (I'm using ns 2.1b5) Thanks, Alex From christiane@trt6.gov.br Wed Jan 12 11:38:45 2000 From: christiane@trt6.gov.br (Christiane) Date: Wed, 12 Jan 2000 09:38:45 -0200 Subject: MPLS simulation Message-ID: <00a301bf5cf1$95c43200$d90a010a@chris> This is a multi-part message in MIME format. ------=_NextPart_000_00A0_01BF5CE0.D1CF0BA0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi everybody, I am trying to simulate a explicited routed path in ns-2 and I am having = some problems with it. For example, can anyone tell me what does the = message "no target for slot 8" means? Did anybody simulated anything related to this (explicited routed = paths)? Any kind of information will be apreciated. Thanks, Christiane Silva da Purifica=E7=E3o Federal University of Pernambuco - Brazil ------=_NextPart_000_00A0_01BF5CE0.D1CF0BA0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi everybody,
 
I am trying to simulate a explicited routed path in = ns-2 and I=20 am having some problems with it. For example, can anyone tell me what = does the=20 message "no target for slot 8" means?
 
Did anybody simulated anything related to this = (explicited=20 routed paths)?
 
Any kind of information will be = apreciated.
 
Thanks,
Christiane Silva da Purifica=E7=E3o
Federal University of Pernambuco - =20 Brazil
------=_NextPart_000_00A0_01BF5CE0.D1CF0BA0-- From L.Wood@eim.surrey.ac.uk Wed Jan 12 13:05:28 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 12 Jan 2000 13:05:28 +0000 (GMT) Subject: MPLS and list archives was Re: ASk for help In-Reply-To: <20000112063054.74172.qmail@hotmail.com> Message-ID: On Wed, 12 Jan 2000, NingKang Jiang wrote: > I want to develop a MPLS module using ns-2 to do some research work > about MPLS technology. > Do you know someone else is doing the same thing? If so,I want to > discuss with him(her). Read http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9912/0144.html and talk to Nikunj. (that particular archived mail isn't listed in the built thread, author, date or subject indices for December 1999, which looks like a bug in Hypermail 2.0b3. So, the list archive is lossy. I've looked through http://www.landfield.com/hypermail/ and current version appears to be 2b29; I've had good experiences with MHonArc in the past.) L. > Thanks, > > My email: nkjiang@hotmail.com > > Ningkang Jiang > 2000.1.12 PGP From Tarik Alj Wed Jan 12 14:56:38 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 12 Jan 2000 09:56:38 -0500 (EST) Subject: use of fid_ with http Message-ID: <200001121456.JAA19709@someware.INRS-Telecom.UQuebec.CA> Hi all, I've been using the http application (based on code of simple-webcache.tcl) and I would like to know what use is made of the fid_ by the http modules? (I have been using the fid_ values in my own code and obviously it's interfering with http). Is there a way to simulate web traffic with fid_ set to 0? Thanks in advance. Tarik From wbliu@eagle.ee.ntu.edu.tw Wed Jan 12 15:56:13 2000 From: wbliu@eagle.ee.ntu.edu.tw (wbliu) Date: Wed, 12 Jan 2000 23:56:13 +0800 Subject: ELN+snoop Message-ID: <004201bf5d15$8e1c9f00$8c14708c@r7942050> hello I am new about Ns Recentely , I want to add the module about "Snoop +ELN" Is there any code release about it?? thanks From Ziyu.Wang@sophia.inria.fr Wed Jan 12 17:20:24 2000 From: Ziyu.Wang@sophia.inria.fr (Ziyu Wang) Date: Wed, 12 Jan 2000 18:20:24 +0100 Subject: A question. Message-ID: <387CB7D8.E7CD9B2F@sophia.inria.fr> Hello, I'm a newer of ns. When I try to follow your example "A new protocol for ns", recompile the file, it tells me : clcl-1.0b8 -I../otcl-1.0a4 -I../tkbox/include -I../tclbox/include -o ping.o ping.cc ping.cc: In method `void PingAgent::recv(Packet *, Handler *)': ping.cc:93: request for member `addr_' in `hdrip->hdr_ip::src_', which is of non-aggregate type `int' make: *** [ping.o] Error 1 Could you tell me how to do with it? I followed your step strictly. Thanks for your help. Best regards and looking forward you. Wang -- PROJET RODEO Tel:33-4-92387990 INSTITUT NATIONAL DE RECHERCHE Fax:33-4-92387765 EN INFORMATIQUE ET EN AUTOMATIQUE Email: INRIA Sophia Antipolis Ziyu.Wang@sophia.inria.fr 2004 route des Lucioles B.P. 93 Pager: 06902 Sophia Antipolis Cedex http://www-sop.inria.fr/rodeo France From xxe101@cs.usask.ca Wed Jan 12 18:58:15 2000 From: xxe101@cs.usask.ca (Mr. Epsilon) Date: Wed, 12 Jan 2000 12:58:15 -0600 (CST) Subject: wireless questions Message-ID: Hello ns users, Is there anybody who has run all the wireless tcl scripts from Marc Greis's tutorial without any problems? I could run only simple-wireless.tcl and wireless1.tcl. For the rest, I can get the .tr and .nam files, but when I run nam I get core dumped. (I have the daily snapshot) Also I want to know the specification for the .tr files. Can we get # of packets lost,tcp and ip throughput, retransmissions from that output file? Thanks... Raja From haoboy@isi.edu Wed Jan 12 19:16:08 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 12 Jan 2000 11:16:08 -0800 (PST) Subject: use of fid_ with http In-Reply-To: <200001121456.JAA19709@someware.INRS-Telecom.UQuebec.CA> Message-ID: It's used to identify which node the server/client/cache resides on. If you want web traffic only, you can use WebTraf which does not require the server/client/cache stuff. See tcl/ex/large-scale-web-traffic.tcl for an example. - Haobo On Wed, 12 Jan 2000, Tarik Alj wrote: > Hi all, > > I've been using the http application (based on code of > simple-webcache.tcl) and I would like to know what use is made of the fid_ by > the http modules? (I have been using the fid_ values in my own code and > obviously it's interfering with http). Is there a way to simulate web traffic > with fid_ set to 0? > > Thanks in advance. > > Tarik > > > From hgossain@ececs.uc.edu Wed Jan 12 19:35:36 2000 From: hgossain@ececs.uc.edu (Hrishikesh Gossain) Date: Wed, 12 Jan 2000 14:35:36 -0500 (EST) Subject: Mobile Multicasting! Message-ID: hi all, i want to know if someone has done work on Mobile Multicasting????? Please let me know if there is some code regarding mobile multicasting in NS. Regards, Hrishi ---------------------------------------------------------------------- Hrishikesh Gossain email: hgossain@ececs.uc.edu Phone No. : (513)556-3437 (Lab.) Research Assistant, Department of ECECS, UC ---------------------------------------------------------------------- From Tarik Alj Wed Jan 12 19:35:55 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 12 Jan 2000 14:35:55 -0500 (EST) Subject: use of fid_ with http Message-ID: <200001121935.OAA21263@someware.INRS-Telecom.UQuebec.CA> Thanks for your fast answer. I thought id_ (from Node) was used for this purp. That's what the code appears to do: http-agent.tcl: Http intsproc init { ns node } {... $self set id_ [$node_ id] ...} . >From the code in http-agent.tcl, http-server.tcl, http-client.tcl, http-cache.tcl, I found out that fid_ was only used with Htpp:getfid in Http::connect and http/Server::alloc-connection. So I've overrided Http::getfid so as to always set fid_ = 0. It seems to work fine for what I need (bursty, exponential traffic). Now I would like to know if what I've done can have a *desastrous* impact on the traffic, if I can reasonnably use it; or if I should switch to large-scale-web-traffic.tcl. > Date: Wed, 12 Jan 2000 11:16:08 -0800 (PST) > From: Haobo Yu > To: Tarik Alj > cc: ns-users@mash.cs.berkeley.edu > Subject: Re: use of fid_ with http > MIME-Version: 1.0 > > It's used to identify which node the server/client/cache resides on. If > you want web traffic only, you can use WebTraf which does not require the > server/client/cache stuff. See tcl/ex/large-scale-web-traffic.tcl for an > example. > > - Haobo > > On Wed, 12 Jan 2000, Tarik Alj wrote: > > > Hi all, > > > > I've been using the http application (based on code of > > simple-webcache.tcl) and I would like to know what use is made of the fid_ by > > the http modules? (I have been using the fid_ values in my own code and > > obviously it's interfering with http). Is there a way to simulate web traffic > > with fid_ set to 0? > > > > Thanks in advance. > > > > Tarik > > > > > > Tarik From haoboy@isi.edu Wed Jan 12 20:31:08 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 12 Jan 2000 12:31:08 -0800 (PST) Subject: use of fid_ with http In-Reply-To: <200001121935.OAA21263@someware.INRS-Telecom.UQuebec.CA> Message-ID: > I thought id_ (from Node) was used for this purp. That's what the code appears > to do: Sorry I didn't explain it clearly. Yes id_ is used for that purpose. fid_ is used to separate every connection that a server/etc makes so that every tcp connection has its own fid_. > So I've overrided Http::getfid so as to always set fid_ = 0. It seems to work > fine for what I need (bursty, exponential traffic). You are right that settting fid_ to 0 does not affect the traffic at all. > Now I would like to know if what I've done can have a *desastrous* impact on the > traffic, if I can reasonnably use it; or if I should switch to > large-scale-web-traffic.tcl. No it has no effect on the traffic. - Haobo From cmc@di.ufpe.br Wed Jan 12 21:12:05 2000 From: cmc@di.ufpe.br (Carlos de Morais Cordeiro) Date: Wed, 12 Jan 2000 19:12:05 -0200 (EDT) Subject: Mobile Multicasting! In-Reply-To: Message-ID: Hi Gossain, We have implemented a protocol called RM2 (Reliable Mobile Multicast) using NS-2. I have some papers which I can make available to you. As for the source code of RM2, I will make it available through a home page in a few weeks. Besides that, I am documenting the implementation so anyone can reuse/modify it. Carlos. On Wed, 12 Jan 2000, Hrishikesh Gossain wrote: > > hi all, > > i want to know if someone has done work on Mobile Multicasting????? > > Please let me know if there is some code regarding mobile multicasting in > NS. > > Regards, > Hrishi > > ---------------------------------------------------------------------- > Hrishikesh Gossain email: hgossain@ececs.uc.edu > Phone No. : (513)556-3437 (Lab.) > Research Assistant, Department of ECECS, UC > ---------------------------------------------------------------------- > > From George.Khoury@era.ericsson.se Wed Jan 12 21:22:05 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Wed, 12 Jan 2000 22:22:05 +0100 Subject: Q_RED Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE004@esealnt190> Dears ns-users, I have a question about RED queue. Could anybody explain what thresh_ and maxthresh_?? I have read on the web that thresh_ = The minimum threshhold for the average queue size in packets!! could anybody explain a bit more!!! thanks in advance George From salehi@isi.edu Wed Jan 12 21:50:23 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 12 Jan 2000 13:50:23 -0800 (PST) Subject: header size in ns In-Reply-To: References: Message-ID: <200001122150.NAA12155@joy.isi.edu> Yan, Are you using Solaris? We've had some problems with byte alignment under Solaris. See if you have the same problem with other OSs and let us know. Nader >>> "Yan" == Yan Yu writes: Yan> Hello, all I am doing simulation using DSR in ns-2. I am trying Yan> to add some extra info in the source route header. so I also Yan> modify th hdr_sr::size() to count the extra filed in sr_addr. BUt Yan> the simulation behaves wierd after I made those change. If I only Yan> change( add a field to) the sr_addr which is included in hdr_sr, Yan> but did not change hdr_size(), which I think it is not right, but Yan> the simulation seems working ok. So I am wondering what change I Yan> need to make when I add some field in packet header, Yan> i.e. hdr_sr. Do I need to change hdr_sr::size()? Is it to count Yan> the size of the route header? Yan> Your any comments would be very helpful and appreciated. Thanks Yan> in advance! yan From engp9075@nus.edu.sg Thu Jan 13 03:05:05 2000 From: engp9075@nus.edu.sg (engp9075@nus.edu.sg) Date: Thu, 13 Jan 2000 11:05:05 +0800 Subject: MPLS simulation References: <00a301bf5cf1$95c43200$d90a010a@chris> Message-ID: <387D40E1.26929046@nus.edu.sg> --------------FF1F194897B375754A25FA7B Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit Hi Christiane, Please read the part about the classifier of ns Notes. "target" here is the object that should receive the packets from that slot. "no target for slot 8" means that the classifier sends the packets to the slot 8 according to some rule, but there is no target for slot 8. For example, the port classifier will send the packets to slot 8 if the port_ is 8, but if there is no link or agent attached at slot 8, the error will occour. Usually, this error in not caused by classifier itself. You'd better check your code carefully at first. Liqun Christiane wrote: > Hi everybody, I am trying to simulate a explicited routed path in > ns-2 and I am having some problems with it. For example, can anyone > tell me what does the message "no target for slot 8" means? Did > anybody simulated anything related to this (explicited routed > paths)? Any kind of information will be apreciated. Thanks,Christiane > Silva da Purifica玢oFederal University of Pernambuco - Brazil --------------FF1F194897B375754A25FA7B Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi Christiane,

Please read the part about the classifier of ns Notes.
"target" here is the object that should receive the packets from that slot.
"no target for slot 8" means that  the classifier sends the packets to the slot
8 according to some rule, but there is no target for slot 8.
For example, the port classifier will send the packets to slot 8 if the port_ is 8,
but if there is no link or agent attached at slot 8, the error will occour.
Usually, this error in not caused by classifier itself. You'd better check your
code carefully at first.

Liqun

Christiane wrote:

 Hi everybody, I am trying to simulate a explicited routed path in ns-2 and I am having some problems with it. For example, can anyone tell me what does the message "no target for slot 8" means? Did anybody simulated anything related to this (explicited routed paths)? Any kind of information will be apreciated. Thanks,Christiane Silva da PurificaçãoFederal University of Pernambuco -  Brazil
--------------FF1F194897B375754A25FA7B-- From Marta.Bordes@CSELT.IT Thu Jan 13 09:45:42 2000 From: Marta.Bordes@CSELT.IT (Bordes Marta) Date: Thu, 13 Jan 2000 10:45:42 +0100 Subject: Cellular IP Message-ID: <9187FF572943D211B28100805FC130FC0169E813@xrr1.cselt.it> Hello everybody, I would like to know if exists an implementation of Cellular IP for ns. Thank you in advance, Marta Bordes. CSELT - Centro Studi e Laboratori Telecomunicazioni Via G. Reiss Romoli 274 10148 TORINO- ITALY Email: marta.bordes@cselt.it From mklun@gte.com Thu Jan 13 16:44:21 2000 From: mklun@gte.com (Miroslav I. Klun) Date: Thu, 13 Jan 2000 11:44:21 -0500 Subject: Random variables question Message-ID: Hi there, I was trying to set a simulation where several CBR sources (video_1, etc.) start at random times, within a specified time interval (say [0.5, 1.0]). Following ns-2 notes & documentation, I thought defining a random variable $u by set u [new RandomVariable/Uniform] $u set min_ 0.5 $u set max_ 1.0 $u value which would be then invoked with likes of $ns at $u "$video_1 start" $ns at 10 "$video_1 stop" etc. But what is actually happening is that all these sources then start at exactly 0.0, obviously a default. Trying to see the value $u itself, by placing puts $u results only in symbol _o13 . I am obviously missing something here. Could anyone please help? Thank you. Miroslav ================================ Dr. Miroslav I. Klun Network Infrastructure Laboratory GTE Laboratories Inc. Waltham, MA 02451 (781) 466-3830 mklun@gte.com From L.Wood@eim.surrey.ac.uk Thu Jan 13 16:48:35 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Thu, 13 Jan 2000 16:48:35 +0000 (GMT) Subject: Random variables question In-Reply-To: Message-ID: $ns at [$u value] "$video_1 start" and puts [$u value] should work. To get a random number out of your RandomVariable/Uniform object, you have to invoke its value method... cheers, L. PGP On Thu, 13 Jan 2000, Miroslav I. Klun wrote: > I was trying to set a simulation where several CBR sources (video_1, etc.) > start at random times, within a specified time interval (say [0.5, 1.0]). > Following ns-2 notes & documentation, I thought defining a random variable > $u by > > set u [new RandomVariable/Uniform] > $u set min_ 0.5 > $u set max_ 1.0 > $u value > > which would be then invoked with likes of > > $ns at $u "$video_1 start" > $ns at 10 "$video_1 stop" > > etc. > > But what is actually happening is that all these sources then > start at exactly 0.0, obviously a default. > > > Trying to see the value $u itself, by placing > > puts $u > > results only in symbol _o13 . > > I am obviously missing something here. Could anyone please help? From Martin Koyabe Thu Jan 13 02:22:26 2000 From: Martin Koyabe (Martin Koyabe) Date: Thu, 13 Jan 2000 02:22:26 +0000 (GMT) Subject: Satellite - Link delay Message-ID: <200001130222.CAA01088@churchward.erg.abdn.ac.uk> Hi All, Are there any of you who have similated satellite delays and bit-errors using NS ? I am interested in simulation, errors, looses (due to bit-errors) and significant asymmetrics for large users in a one-hop scenario for a geo-satellite case !! I have the following topology:- create network topology like this: (3) Receiver 1 / / Source (0)--------1 Satellite \ \ (2) Receiver 2 Would the a code like this make sense:- $ns simplex-link $nodes_(0) $nodes_(1) 4Mb 125ms $ns simplex-link $nodes_(1) $nodes_(0) 1Mb 125ms $ns simplex-link $nodes_(1) $nodes_(2) 4Mb 125ms $ns simplex-link $nodes_(1) $nodes_(3) 4Mb 125ms $ns simplex-link $nodes_(2) $nodes_(1) 1Mb 125ms $ns simplex-link $nodes_(3) $nodes_(1) 1Mb 125ms I'll appreciate any comments/suggestions/leads !! -- Martin --------------------------------------------------------------------- Martin L.W.D Koyabe e-mail: koyabe@erg.abdn.ac.uk Electronics Research Group Fraser Noble Building King's College, AB24 3UE (Off) +44-01224-272813 Aberdeen, UK (Mob) +44-07881-610825 From mallman@grc.nasa.gov Thu Jan 13 17:14:37 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Thu, 13 Jan 2000 12:14:37 -0500 Subject: problem in newtimer() (?) Message-ID: <200001131714.MAA07177@lombok-fi.lerc.nasa.gov> I think I have run across a bug in ns' one-way TCP implementation. I can only really produce this effect in a big simulation with lots of stuff going on, so I haven't been able to fully nail down what should be happening. But, it goes something like this... We send a load of packets in slow start and they are nearly all lost. Recovery happens naturally for a little while. Then, things just end... t_seqno_ is set to 406. I get an ACK for 406. This ACK doesn't let me transmit anything (using SACK versionof fast recovery). However, according to TcpAgent::newtimer() I cancel the RTO timer, even though I have sent many more than 406 packets. I get a duplicate ACK, which also doesn't let me transmit anything. Now, the transfer is done. It is hung because there is no rexmt timer, I think. So, I made a patch that seems to help the situation. Here is my version of the newtimer() function. void TcpAgent::newtimer(Packet* pkt) { hdr_tcp *tcph = hdr_tcp::access(pkt); #ifdef MALLMAN if ((t_seqno_ > tcph->seqno()) || (tcph->seqno() < maxseq_)) #else if (t_seqno_ > tcph->seqno()) #endif MALLMAN set_rtx_timer(); else cancel_rtx_timer(); } In other words, we don't cancel the RTO timer if we have not received an ACK for the last thing we have sent. This seems to have fixed the problem. However, I am not sure if that introduces other problems. I'd appreciate some feedback on what this could effect. Also, I can provide the packet traces I have if someone wants to dig into this a little more. allman From wbliu@eagle.ee.ntu.edu.tw Thu Jan 13 17:51:57 2000 From: wbliu@eagle.ee.ntu.edu.tw (wbliu) Date: Fri, 14 Jan 2000 01:51:57 +0800 Subject: ELN+snoop Message-ID: <006501bf5dee$e3835d60$8514708c@r7942050> hello I am new about Ns Recentely , I want to add the module about "Snoop +ELN" Is there any code release about it?? or could you teach me how to add it?? thanks From javierg@ctr.columbia.edu Thu Jan 13 19:06:29 2000 From: javierg@ctr.columbia.edu (Javier Gomez-Castellanos) Date: Thu, 13 Jan 2000 14:06:29 -0500 (EST) Subject: Cellular IP (fwd) In-Reply-To: <387E43D5.B1DC3CB3@comet.columbia.edu> Message-ID: yes, there is an implementation of cellular IP in NS, we hope to make it available soon (needs some cleaning). We have a paper related to the simulation and the source code of the testbed is available at our web site. http://comet.ctr.columbia.edu/cellularip/ --Javier ----------------------------------------- COMET Group Columbia University, New York. > > > > ---------- Forwarded message ---------- > > Date: Thu, 13 Jan 2000 10:45:42 +0100 > > From: Bordes Marta > > To: "'ns-users@mash.cs.berkeley.edu> > > > Hello everybody, > > > > I would like to know if exists an implementation of Cellular IP for ns. > > > > Thank you in advance, > > > > Marta Bordes. > > > > CSELT - Centro Studi e Laboratori Telecomunicazioni > > Via G. Reiss Romoli 274 > > 10148 TORINO- ITALY > > > > Email: marta.bordes@cselt.it > From lli@shtdu.edu.cn Thu Jan 13 21:10:07 2000 From: lli@shtdu.edu.cn (Li Lei) Date: Fri, 14 Jan 2000 5:10:7 +0800 Subject: question of red trace Message-ID: <200001132109.FAA13130@ suda.edu.cn> Dear all, I used ns2.1b5 to simulate RED gateway and I traced linterm value using test-suite-red-v1a.tcl. If I set a linterm a new value different to the default value(10), such as 4, I was very surprized to find that the linterm was changing during the simulation, sometimes is 4, sometimes is 10. When I traced the other parameters, the same thing happened, they were all changed to the default value now and then. Why? Thanks a lot. still From Martin Koyabe Thu Jan 13 06:46:59 2000 From: Martin Koyabe (Martin Koyabe) Date: Thu, 13 Jan 2000 06:46:59 +0000 (GMT) Subject: ns-2.1b6 Compilation Message-ID: <200001130646.GAA02441@churchward.erg.abdn.ac.uk> Hi All, Could someone tell me what this error means and possible solutions when I try to update my ns-2.1b5 installation to the current 'snapshot' ? ---------- Koyabe @ churchward 70 > make c++ -c -O2 -DTCP_DELAY_BIND_ALL -DNO_TK -DTCLCL_CLASSINSTVAR -DNDEBUG -mv8 -msupersparc -D__svr4__ -DUSE_SHM -DHAVE_LIBTCLCL1_0B8 -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_0A4 -DHAVE_OTCL_H -DHAVE_LIBTK8_0 -DHAVE_TK_H -DHAVE_LIBTCL8_0 -DHAVE_TCL_H -DHAVE_CONFIG_H -I. -I/home/churchward/koyabe/ns/tclcl-1.0b8 -I/home/churchward/koyabe/ns/otcl-1.0a4 -I/home/churchward/koyabe/ns/tk8.0/generic -I/home/churchward/koyabe/ns/tcl8.0/generic -o random.o random.cc In file included from /home/churchward/koyabe/ns/tclcl-1.0b8/tclcl.h:47, from config.h:60, from random.cc:40: /home/churchward/koyabe/ns/tclcl-1.0b8/tclcl-config.h:37: warning: `HAVE_INT64' redefined autoconf.h:46: warning: this is the location of the previous definition --------- Regards -- Martin From alhussein_abouzeid@hotmail.com Thu Jan 13 21:34:04 2000 From: alhussein_abouzeid@hotmail.com (Alhussein Abouzeid) Date: Thu, 13 Jan 2000 21:34:04 GMT Subject: problem in newtimer() (?) Message-ID: <20000113213404.57605.qmail@hotmail.com> This problem sounds very much like one I had earlier last year. In my case, if a really long sequence of packet transmissions are lost (whether in slows start or not), something happens that doesn't allow TCP to transmit again, even if I increase the simulation length to exceed 64*RTT (the value of the time-out period due to binary exponential back-off). I think the probelm might be in there, in the BEBO algorithm, but I didn't have the time to really go after this possible bug. Hope this helps, Hussein. >From: Mark Allman >Reply-To: mallman@grc.nasa.gov >To: ns-users@mash.cs.berkeley.edu >Subject: problem in newtimer() (?) >Date: Thu, 13 Jan 2000 12:14:37 -0500 > >I think I have run across a bug in ns' one-way TCP implementation. >I can only really produce this effect in a big simulation with lots >of stuff going on, so I haven't been able to fully nail down what >should be happening. But, it goes something like this... > >We send a load of packets in slow start and they are nearly all >lost. Recovery happens naturally for a little while. Then, things >just end... t_seqno_ is set to 406. I get an ACK for 406. This >ACK doesn't let me transmit anything (using SACK versionof fast >recovery). However, according to TcpAgent::newtimer() I cancel the >RTO timer, even though I have sent many more than 406 packets. I >get a duplicate ACK, which also doesn't let me transmit anything. >Now, the transfer is done. It is hung because there is no rexmt >timer, I think. So, I made a patch that seems to help the >situation. Here is my version of the newtimer() function. > >void TcpAgent::newtimer(Packet* pkt) >{ > hdr_tcp *tcph = hdr_tcp::access(pkt); >#ifdef MALLMAN > if ((t_seqno_ > tcph->seqno()) || (tcph->seqno() < maxseq_)) >#else > if (t_seqno_ > tcph->seqno()) >#endif MALLMAN > set_rtx_timer(); > else > cancel_rtx_timer(); >} > >In other words, we don't cancel the RTO timer if we have not >received an ACK for the last thing we have sent. > >This seems to have fixed the problem. However, I am not sure if >that introduces other problems. I'd appreciate some feedback on >what this could effect. Also, I can provide the packet traces I >have if someone wants to dig into this a little more. > >allman ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From Martin Koyabe Thu Jan 13 07:06:25 2000 From: Martin Koyabe (Martin Koyabe) Date: Thu, 13 Jan 2000 07:06:25 +0000 (GMT) Subject: Mailing List Server Date Message-ID: <200001130706.HAA02472@churchward.erg.abdn.ac.uk> Hi, Thought this might be helpful to point out. The mailing list server for ns-users hosting Jan-2000 archieves has a wrong date !! Please update it !! check !! http://www-mash.cs.berkeley.edu/dist/archive/ns-users/0001/subject.html Starting: Fri 02 Jan 1970 - 19:37:51 PST Ending: Tue 13 Jan 1970 - 01:48:18 PST Messages: 91 -- Martin From johnh@isi.edu Thu Jan 13 23:35:14 2000 From: johnh@isi.edu (John Heidemann) Date: Thu, 13 Jan 2000 15:35:14 -0800 Subject: Mailing List Server Date In-Reply-To: Message from Martin Koyabe of "Thu, 13 Jan 2000 07:06:25 GMT." <200001130706.HAA02472@churchward.erg.abdn.ac.uk> Message-ID: <200001132335.PAA17763@dash.isi.edu> >Hi, > >Thought this might be helpful to point out. The mailing list server for ns-users >hosting Jan-2000 archieves has a wrong date !! Please update it !! > >check !! > >http://www-mash.cs.berkeley.edu/dist/archive/ns-users/0001/subject.html > >Starting: Fri 02 Jan 1970 - 19:37:51 PST >Ending: Tue 13 Jan 1970 - 01:48:18 PST >Messages: 91 Thanks for the note. We'll try to update this from our bunker as soon as we have power restored :-) -John Heidemann From salehi@isi.edu Fri Jan 14 02:05:58 2000 From: salehi@isi.edu (Nader Salehi) Date: Thu, 13 Jan 2000 18:05:58 -0800 (PST) Subject: ns-2.1b6 Compilation In-Reply-To: <200001130646.GAA02441@churchward.erg.abdn.ac.uk> References: <200001130646.GAA02441@churchward.erg.abdn.ac.uk> Message-ID: <200001140205.SAA19206@joy.isi.edu> You need to download the most recent snapshot of tclcl, recofingure, and build it. Nader >>> "Martin" == Martin Koyabe writes: Martin> Hi All, Could someone tell me what this error means and Martin> possible solutions when I try to update my ns-2.1b5 Martin> installation to the current 'snapshot' ? Martin> ---------- Koyabe @ churchward 70 > make c++ -c -O2 Martin> -DTCP_DELAY_BIND_ALL -DNO_TK -DTCLCL_CLASSINSTVAR -DNDEBUG Martin> -mv8 -msupersparc -D__svr4__ -DUSE_SHM -DHAVE_LIBTCLCL1_0B8 Martin> -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_0A4 -DHAVE_OTCL_H Martin> -DHAVE_LIBTK8_0 -DHAVE_TK_H -DHAVE_LIBTCL8_0 -DHAVE_TCL_H Martin> -DHAVE_CONFIG_H -I. -I/home/churchward/koyabe/ns/tclcl-1.0b8 Martin> -I/home/churchward/koyabe/ns/otcl-1.0a4 Martin> -I/home/churchward/koyabe/ns/tk8.0/generic Martin> -I/home/churchward/koyabe/ns/tcl8.0/generic -o random.o Martin> random.cc In file included from Martin> /home/churchward/koyabe/ns/tclcl-1.0b8/tclcl.h:47, from Martin> config.h:60, from random.cc:40: Martin> /home/churchward/koyabe/ns/tclcl-1.0b8/tclcl-config.h:37: Martin> warning: `HAVE_INT64' redefined autoconf.h:46: warning: this Martin> is the location of the previous definition Martin> --------- Martin> Regards Martin> -- Martin From rosblom@yahoo.com Fri Jan 14 02:56:22 2000 From: rosblom@yahoo.com (=?iso-8859-1?q?Andreas=20Rosblom?=) Date: Thu, 13 Jan 2000 18:56:22 -0800 (PST) Subject: Which version of RedHat Linux ??? Message-ID: <20000114025622.24324.qmail@web1203.mail.yahoo.com> Does anybody know which version of Linux RedHat that ns-2b5 works on? Please, state the latest one. Thanks! - Andreas __________________________________________________ Do You Yahoo!? Talk to your friends online with Yahoo! Messenger. http://im.yahoo.com From hgossain@ececs.uc.edu Fri Jan 14 03:25:18 2000 From: hgossain@ececs.uc.edu (Hrishikesh Gossain) Date: Thu, 13 Jan 2000 22:25:18 -0500 (EST) Subject: Core Dump in wireless3.tcl Message-ID: hi ana, i too faced the same problem. But "diepchi" suggested me to change the namtrace-all-wireless to namtrace-all and comment out $ns_ initial_node_pos $node_($i) 20 . In that way although nam runs, but we do not have animation of mobile part. In a sense it will show only wired part. I am also waiting for some code/application which shows both wired and wireless animation. I don't know, but i just wonder if this is something related to the fact of developing wired and wireless parts by two different groups?????? thanks, Hrishi ---------------------------------------------------------------------- Hrishikesh Gossain email: hgossain@ececs.uc.edu Phone No. : (513)556-3437 (Lab.) Research Assistant, Department of ECECS, UC ---------------------------------------------------------------------- Hi everyone, I have been trying to run wireless3.tcl from Marc Greil's tutorial and when I call nam to visualize the topology I obtain a Segmentation Fault or Core Dump. Did anyone have the same problem? Thanks in advance. Cheers, Ana ________________________________ Ana Delgado Centre for Telecommunication Research King's College London London England ana_delgado77@hotmail.com From charles.a.nunez@lmco.com Fri Jan 14 03:36:22 2000 From: charles.a.nunez@lmco.com (Nunez, Charles A) Date: Thu, 13 Jan 2000 19:36:22 -0800 Subject: Which version of RedHat Linux ??? Message-ID: <079B626B05A0D3118B1000508B0EA5E9289930@emss04m05.ems.lmco.com> It works for me on Version 6.0. Chuck > -----Original Message----- > From: Andreas Rosblom [SMTP:rosblom@yahoo.com] > Sent: Thursday, January 13, 2000 9:56 PM > To: ns-users@mash.CS.Berkeley.EDU > Subject: Which version of RedHat Linux ??? > > Does anybody know which version of Linux RedHat that > ns-2b5 works on? Please, state the latest one. Thanks! > > - Andreas > __________________________________________________ > Do You Yahoo!? > Talk to your friends online with Yahoo! Messenger. > http://im.yahoo.com From Martin Koyabe Thu Jan 13 22:18:28 2000 From: Martin Koyabe (Martin Koyabe) Date: Thu, 13 Jan 2000 22:18:28 +0000 (GMT) Subject: ns-2.16b compilation Message-ID: <200001132218.WAA18072@churchward.erg.abdn.ac.uk> Hi all, I seems to have got all the compilations working for (tk8.0.4,tcl8.0.4), snapshots of tclcl-1.0b9, buts when I compile ns-2.1b6 (todays -14/1/2000 snaptshot), I get the following error !! ------ Undefined first referenced symbol in file TclObject::delay_bind_dispatch(char const *, char const *)rng.o ld: fatal: Symbol referencing errors. No output written to ns collect2: ld returned 1 exit status make: *** [ns] Error 1 ------- Any one knows what the error is !! -- Martin From puginifabio@tiscalinet.it Fri Jan 14 14:10:31 2000 From: puginifabio@tiscalinet.it (puginifabio) Date: Fri, 14 Jan 2000 06:10:31 -0800 (PST) Subject: Dual Leaky Bucket Message-ID: <387f2d6e38d70396@twingo.tiscalinet.it> (added by twingo.tiscalinet.it) Hi I'm an italian student and I'm facing ns for the first time. Does any Dual Leaky Bucket already exist? thanks. -- TiscaliFreeNet, libero accesso ad Internet. http://www.tiscalinet.it From Tan H P" This is a multi-part message in MIME format. ------=_NextPart_000_000A_01BF5EDC.9F0FEBA0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi all,=20 I am trying to run a wireless-cum-wired simulation with TCP traffic. I = tried to do a fragmentation at the link layer, and I found that as more = and more packets are generated, some packets are lost in the IFQ. Has = anyone experienced this before? Is there a memory limitation of ns?=20 I did the fragmentation as follows : Many copies of each LL frame is = generated and sent down to MAC. At the receiving MAC, these packets are = collected to form a frame, and then sent to the Link Layer.=20 Thanks for any reply.=20 Regards HP Tan ------=_NextPart_000_000A_01BF5EDC.9F0FEBA0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi all,
 
I am trying to run a wireless-cum-wired simulation = with TCP=20 traffic. I tried to do a fragmentation at the link layer, and I found = that as=20 more and more packets are generated, some packets are lost in the IFQ. = Has=20 anyone experienced this before? Is there a memory limitation of ns?=20
 
I did the fragmentation as follows : Many copies of = each LL=20 frame is generated and sent down to MAC. At the receiving MAC, these = packets are=20 collected to form a frame, and then sent to the Link Layer. =
 
Thanks for any reply.

Regards
HP Tan
------=_NextPart_000_000A_01BF5EDC.9F0FEBA0-- From klan@usc.edu Fri Jan 14 16:01:54 2000 From: klan@usc.edu (Kun-chan Lan) Date: Fri, 14 Jan 2000 08:01:54 -0800 Subject: ns-2.16b compilation Message-ID: <001701bf5ea8$ae416060$0100a8c0@trojan.usc.edu> please update your tclcl to the current snapshot . kun-chan -----Original Message----- From: Martin Koyabe To: ns-users@mash.CS.Berkeley.EDU Date: Friday, January 14, 2000 5:03 AM Subject: ns-2.16b compilation >Hi all, > >I seems to have got all the compilations working for (tk8.0.4,tcl8.0.4), >snapshots of tclcl-1.0b9, buts when I compile ns-2.1b6 (todays -14/1/2000 >snaptshot), I get the following error !! > >------ >Undefined first referenced > symbol in file >TclObject::delay_bind_dispatch(char const *, char const *)rng.o >ld: fatal: Symbol referencing errors. No output written to ns >collect2: ld returned 1 exit status >make: *** [ns] Error 1 >------- > >Any one knows what the error is !! > >-- Martin > > From jamester@dacl3.snu.ac.kr Fri Jan 14 15:52:45 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Sat, 15 Jan 2000 00:52:45 +0900 Subject: Replicator References: Message-ID: <00011501150500.32743@jamester.snu.ac.kr> Hello All. I gave up using get-mcast-tree to find out intermediate nodes of a multicast tree. Instead, after all the receivers join group and after sufficient amount of time, I call " set r [$node($i) getRepByGroup $group" for all nodes and consider the node is in the tree if { $r != "" && [$r is-active]}. This works fine only one time. However, I encounter problem when I do following # All the agents are created before this loop set time 0 for { set i 0 } { $i < [lindex $argv 1[ } { incr i } { $ns at $time --- get random source id : src --- attach $node($src) source_agent incr time $ns at $time --- get random receiver ids : rid --- attach $node($rid) receiver_agents incr time $ns at $time --- source agent start incr time $ns at $time --- receivers join group set time [expr $time + 10] $ns at $time foreach n [array name node] { set r [$n getRepByGroup $group] if { $r != "" && [$r is-active] } { #intermediate node found lappend inter_nodes $n } } --- gather statistics and compute average, worst and best cases incr time $ns at $time --- stop source agent --- all receivers leave the group --- detach all the agents from nodes --- set inter_nodes "" incr time } When the loop runs for the second time, some nodes contain 2 active replicators for the group and 3 active replicators for the third time and so on. When there are more than one replicators [$r is-active] generates an error which I can solve. My question is that when receivers leave the group, shouldn't replicator for the group become inactive or removed? In addition, it seems like multiple active replicators for a group per a node really slows down the simulation process as well. Thanks in advance -James Kwon From George.Khoury@era.ericsson.se Fri Jan 14 17:30:05 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Fri, 14 Jan 2000 18:30:05 +0100 Subject: Simulation Error Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE008@esealnt190> Dear ns-users, I have a small question: I'm trying to simulate a "big network " with few nodes and many sources and when I try to simulate over 250 sec. I got massages that the simulator is going back in time. I wonder if this a problem in the computer ( a short of memory ) or if this is a ns problem!! Any suggestions would be appreciated!! Thanks in advance From huiwenli@nortelnetworks.com Fri Jan 14 19:38:27 2000 From: huiwenli@nortelnetworks.com (Huiwen Li) Date: Fri, 14 Jan 2000 11:38:27 -0800 Subject: NS installation and configuration question Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF5EC6.ED2D4208 Content-Type: text/plain; charset="iso-8859-1" Hi, all, I just noticed this problem although I have installed ns for several months. In my machine, there is an old version of Tcl and Tk, which is version7.4. At that time I installed the allinone package. So I didn't do much configuration. I found although install script was configured to use the right version under /ns-allinone-2.1b5, under the ns-2.1b5 directory, configure file was using the older version, because prefix is detected to be /usr/local/bin which includes the older version. I am wondering which one is ns exactly using? Should I recompile ns separately by configure it with ./configure --with-tcl=.....? Or should I define some path in my .cshrc file? When I type "which tclsh" it showed me both these versions with the newer one first, but when I type "info tclversion" it showed me 8.0. I am really confused. How to explain the Makefile for ns-2.1b5 then? It is as follows: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ # # Various configurable paths (remember to edit Makefile.in, not Makefile) # # Top level hierarchy prefix = /usr/local # Pathname of directory to install the binary BINDEST = /usr/local/bin # Pathname of directory to install the man page MANDEST = /usr/local/man BLANK = # make a blank space. DO NOT add anything to this line # The following will be redefined under Windows (see WIN32 lable below) CC = gcc CPP = c++ LINK = $(CPP) MKDEP = ./conf/mkdep TCLSH = /usr/local/bin/tclsh TCL2C = ../tclcl-1.0b8/tcl2c++ AR = ar rc $(BLANK) RANLIB = ranlib INSTALL = /mod/gnu/bin/install -c LN = ln TEST = test RM = rm -f PERL = /usr/local/bin/perl CCOPT = STATIC = LDFLAGS = $(STATIC) LDOUT = -o $(BLANK) DEFINE = -DNO_TK -DTCLCL_CLASSINSTVAR -DNDEBUG -mv8 -msupersparc -D__svr4__ -DUSE_SHM -DHAVE_LIBTCLCL1_0B8 -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_0A4 -DHAVE_OTCL_H -DHAVE_LIBTK8_0 -DHAVE_TK_H -DHAVE_LIBTCL8_0 -DHAVE_TCL_H -DHAVE_CONFIG_H INCLUDES = \ -I. \ -I../tclcl-1.0b8 -I../otcl-1.0a4 -I/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/generic -I../tcl8.0.4/generic LIB = \ -R../tclcl-1.0b8 -L../tclcl-1.0b8 -ltclcl -R../otcl-1.0a4 -L../otcl-1.0a4 -lotcl -R/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix -L/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix -ltk8.0 -R../tcl8.0.4/unix -L../tcl8.0.4/unix -ltcl8.0 \ -lXext -lX11 \ -lsocket -lnsl -lintl -ldl -ldl \ -lm -ldl # -L${exec_prefix}/lib \ CFLAGS = $(CCOPT) $(DEFINE) ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Thank you for your time. I really appreciate any clarification. Huiwen ------_=_NextPart_001_01BF5EC6.ED2D4208 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable NS installation and configuration question

Hi, all,

I just noticed this problem although = I have installed ns for several months. In my machine, there is an old = version of Tcl and Tk, which is version7.4. At that time I installed = the allinone package. So I didn't do much configuration. I found = although install script was configured to use the right version under = /ns-allinone-2.1b5, under the ns-2.1b5 directory, configure file was = using the older version, because prefix is detected to be = /usr/local/bin which includes the older version. I am wondering which = one is ns exactly using? Should I recompile ns separately by configure = it with ./configure --with-tcl=3D.....? Or should I define some path in = my .cshrc file? When I type "which tclsh" it showed me both = these versions with the newer one first, but when I type "info = tclversion" it showed me 8.0. I am really confused.

How to explain the Makefile for = ns-2.1b5 then? It is as follows:

    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~=
    #
    # Various configurable paths = (remember to edit Makefile.in, not Makefile)
    #

    # Top level hierarchy
    prefix  =3D = /usr/local
    # Pathname of directory to = install the binary
    BINDEST =3D = /usr/local/bin
    # Pathname of directory to = install the man page
    MANDEST =3D = /usr/local/man

    BLANK   =3D # make a = blank space.  DO NOT add anything to this line

    # The following will be = redefined under Windows (see WIN32 lable below)
    CC      =3D gcc
    CPP     =3D = c++
    LINK    =3D = $(CPP)
    MKDEP   =3D = ./conf/mkdep
    TCLSH   =3D = /usr/local/bin/tclsh
    TCL2C   =3D = ../tclcl-1.0b8/tcl2c++
    AR      =3D ar rc $(BLANK)

    RANLIB  =3D ranlib
    INSTALL =3D = /mod/gnu/bin/install -c
    LN      =3D ln
    TEST    =3D = test
    RM      =3D rm -f
    PERL    =3D = /usr/local/bin/perl


    CCOPT   =3D
    STATIC  =3D
    LDFLAGS =3D $(STATIC)
    LDOUT   =3D -o = $(BLANK)

    DEFINE  =3D -DNO_TK = -DTCLCL_CLASSINSTVAR  -DNDEBUG -mv8 -msupersparc -D__svr4__ = -DUSE_SHM -DHAVE_LIBTCLCL1_0B8 -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_0A4 = -DHAVE_OTCL_H -DHAVE_LIBTK8_0 -DHAVE_TK_H -DHAVE_LIBTCL8_0 -DHAVE_TCL_H&= nbsp; -DHAVE_CONFIG_H

    INCLUDES =3D \
            -I.  \
            -I../tclcl-1.0b8 -I../otcl-1.0a4 = -I/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/generic = -I../tcl8.0.4/generic

    LIB     =3D = \
            -R../tclcl-1.0b8 -L../tclcl-1.0b8 -ltclcl = -R../otcl-1.0a4 -L../otcl-1.0a4 -lotcl = -R/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix = -L/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix -ltk8.0 = -R../tcl8.0.4/unix -L../tcl8.0.4/unix -ltcl8.0  \

            -lXext -lX11 \
             -lsocket -lnsl -lintl -ldl -ldl \
            -lm -ldl
    #       -L${exec_prefix}/lib = \

    CFLAGS  =3D $(CCOPT) = $(DEFINE)
    ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

    Thank you for your time. I really = appreciate any clarification.

    Huiwen

------_=_NextPart_001_01BF5EC6.ED2D4208-- From bks@CS.Berkeley.EDU Fri Jan 14 20:57:23 2000 From: bks@CS.Berkeley.EDU (Brian K. Shiratsuki) Date: Fri, 14 Jan 2000 12:57:23 -0800 Subject: Mailing List Server Date Message-ID: <200001142057.MAA41971@destroyer.CS.Berkeley.EDU> Hypermail now produces correct dates for the January 2000 ns-users archives. -- Brian From haoboy@isi.edu Fri Jan 14 21:31:39 2000 From: haoboy@isi.edu (Haobo Yu) Date: Fri, 14 Jan 2000 13:31:39 -0800 (PST) Subject: NS installation and configuration question In-Reply-To: Message-ID: On FreeBSD tcl/tk are installed under different directories under /usr/local/{include,lib} so ns will not mess up versions. It's unclear what's your case, but you can always check tcl version by typing 'set tcl_version' inside ns. Note ns does not need tk. - Haobo On Fri, 14 Jan 2000, Huiwen Li wrote: > Hi, all, > > I just noticed this problem although I have installed ns for several months. > In my machine, there is an old version of Tcl and Tk, which is version7.4. > At that time I installed the allinone package. So I didn't do much > configuration. I found although install script was configured to use the > right version under /ns-allinone-2.1b5, under the ns-2.1b5 directory, > configure file was using the older version, because prefix is detected to be > /usr/local/bin which includes the older version. I am wondering which one is > ns exactly using? Should I recompile ns separately by configure it with > ./configure --with-tcl=.....? Or should I define some path in my .cshrc > file? When I type "which tclsh" it showed me both these versions with the > newer one first, but when I type "info tclversion" it showed me 8.0. I am > really confused. > > How to explain the Makefile for ns-2.1b5 then? It is as follows: > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > # > # Various configurable paths (remember to edit Makefile.in, not > Makefile) > # > > # Top level hierarchy > prefix = /usr/local > # Pathname of directory to install the binary > BINDEST = /usr/local/bin > # Pathname of directory to install the man page > MANDEST = /usr/local/man > > BLANK = # make a blank space. DO NOT add anything to this line > > # The following will be redefined under Windows (see WIN32 lable > below) > CC = gcc > CPP = c++ > LINK = $(CPP) > MKDEP = ./conf/mkdep > TCLSH = /usr/local/bin/tclsh > TCL2C = ../tclcl-1.0b8/tcl2c++ > AR = ar rc $(BLANK) > > RANLIB = ranlib > INSTALL = /mod/gnu/bin/install -c > LN = ln > TEST = test > RM = rm -f > PERL = /usr/local/bin/perl > > > CCOPT = > STATIC = > LDFLAGS = $(STATIC) > LDOUT = -o $(BLANK) > > DEFINE = -DNO_TK -DTCLCL_CLASSINSTVAR -DNDEBUG -mv8 -msupersparc > -D__svr4__ -DUSE_SHM -DHAVE_LIBTCLCL1_0B8 -DHAVE_TCLCL_H -DHAVE_LIBOTCL1_0A4 > -DHAVE_OTCL_H -DHAVE_LIBTK8_0 -DHAVE_TK_H -DHAVE_LIBTCL8_0 -DHAVE_TCL_H > -DHAVE_CONFIG_H > > INCLUDES = \ > -I. \ > -I../tclcl-1.0b8 -I../otcl-1.0a4 > -I/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/generic > -I../tcl8.0.4/generic > > LIB = \ > -R../tclcl-1.0b8 -L../tclcl-1.0b8 -ltclcl -R../otcl-1.0a4 > -L../otcl-1.0a4 -lotcl > -R/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix > -L/var/PACKAGE/ns-allinone-2.1b5/ns-2.1b5/../tk8.0.4/unix -ltk8.0 > -R../tcl8.0.4/unix -L../tcl8.0.4/unix -ltcl8.0 \ > -lXext -lX11 \ > -lsocket -lnsl -lintl -ldl -ldl \ > -lm -ldl > # -L${exec_prefix}/lib \ > > CFLAGS = $(CCOPT) $(DEFINE) > ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ > > Thank you for your time. I really appreciate any clarification. > > Huiwen > > From kts@server.gigabitwireless.com Fri Jan 14 23:57:19 2000 From: kts@server.gigabitwireless.com (Kim) Date: Fri, 14 Jan 2000 15:57:19 -0800 (PST) Subject: Ques. about ns-2/delaymodel Message-ID: Hi, I am trying to make the delay (prop + txmission) random for a link. Could somebody provide clues how to do this? I can't figure out what delaymodel(.h and .cc) are doing, and what delay(.h and .cc) are doing? The random delay will come from either a trace/profile or from a distribution. Thanks Tai Suk. From floyd@aciri.org Sat Jan 15 00:41:08 2000 From: floyd@aciri.org (Sally Floyd) Date: Fri, 14 Jan 2000 16:41:08 -0800 Subject: problem in newtimer() (?) Message-ID: <200001150041.QAA84636@elk.aciri.org> Mark - >I think I have run across a bug in ns' one-way TCP implementation. ... >It is hung because there is no rexmt >timer, I think. So, I made a patch that seems to help the >situation. Here is my version of the newtimer() function. I wasn't able to (quickly) construct a validation test that exhibits the bug, but it is clearly a bug in the code, and I have put in your fix. (It doesn't make any change in the existing validation tests one way or another.) The existing comments for newtimer() were as follows: /* * Set new retransmission timer if not all outstanding * or available data acked. Otherwise, if a timer is still * outstanding, cancel it. */ but the code in newtimer() did not accurately check if all outstanding data had been acked. In particular, after a retransmit timeout, the variable "t_seqno_" is reset (decreased) to highest_ack_ + 1, but we might have outstanding unacknowledged packets higher than "t_seqno_", up to "maxseq_", that have not been acknowledged. The timeout() code does the check correctly, I believe, by checking "maxseq_" instead of "t_seqno_"... The other functionally-equivalent fix would have been not to cancel the rtx_timer in newtimer(), and to simply wait until it expires in timeout(), where the check about whether to set another timer was already correct... (This fix will not go in the upcoming release, which is unfortunately based on yesterday's code...) - Sally -------------------------------- http://www.aciri.org/floyd/ -------------------------------- From floyd@aciri.org Sat Jan 15 00:57:07 2000 From: floyd@aciri.org (Sally Floyd) Date: Fri, 14 Jan 2000 16:57:07 -0800 Subject: bug report - TCPSink/*, packetSize_, headers and recvBytes() Message-ID: <200001150057.QAA85074@elk.aciri.org> Lloyd - ... >One-way TCPSinks consider packetSize_ (confusingly defined as >size_ rather than packetSize_ in C++, alas) to be the segment size as >well, without taking header overhead into account at any point after >receiving a packet and before passing received segment data from the >window up to recvBytes() for an application to take. Yep, you are right, that is what one-way TCP does, and it is certainly not what one would like. (I will spare you the historical reasons why it all evolved that way...) Unfortunately, I don't have the cycles right now to take it on and change it, but possibly someone else will volunteer. Otherwise it will have to wait until I get to it... - Sally -------------------------------- http://www.aciri.org/floyd/ -------------------------------- ... >I'm blaming this all on calling it size_, you know. Yes, doing something cheap, quick and easy the first time around, as is generally my own inclination, often manages to come back to haunt... From engp9075@nus.edu.sg Sat Jan 15 04:51:11 2000 From: engp9075@nus.edu.sg (engp9075@nus.edu.sg) Date: Sat, 15 Jan 2000 12:51:11 +0800 Subject: Memory problems with ns References: <000d01bf5e99$918b5ca0$0df3fea9@oemcomputer> Message-ID: <387FFCBF.F780B30B@nus.edu.sg> --------------6E9B0AC2F459B192FF4B5A31 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi H P, It seems that you have made some progress. I do not know your simulation configuration. As I know the default length of IFQ is 50 packets. If you have not change it, you can change it to a larger value and be sure that your traffic source sends the packets slower than the MAC layer does. Best regards. Liqun Tan H P wrote: > Hi all, I am trying to run a wireless-cum-wired simulation with TCP > traffic. I tried to do a fragmentation at the link layer, and I found > that as more and more packets are generated, some packets are lost in > the IFQ. Has anyone experienced this before? Is there a memory > limitation of ns? I did the fragmentation as follows : Many copies of > each LL frame is generated and sent down to MAC. At the receiving MAC, > these packets are collected to form a frame, and then sent to the Link > Layer. Thanks for any reply. > > RegardsHP Tan --------------6E9B0AC2F459B192FF4B5A31 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi  H P,
   It seems that you have made some progress.
    I do not know your simulation configuration. As I know the
default length of IFQ is 50 packets. If you have not change it,
you can change it to a larger value and be sure that your traffic
source sends the packets slower than the MAC  layer does.

Best regards.
Liqun
 
 
 

Tan H P wrote:

Hi all, I am trying to run a wireless-cum-wired simulation with TCP traffic. I tried to do a fragmentation at the link layer, and I found that as more and more packets are generated, some packets are lost in the IFQ. Has anyone experienced this before? Is there a memory limitation of ns? I did the fragmentation as follows : Many copies of each LL frame is generated and sent down to MAC. At the receiving MAC, these packets are collected to form a frame, and then sent to the Link Layer. Thanks for any reply.

RegardsHP Tan

--------------6E9B0AC2F459B192FF4B5A31-- From L.Wood@eim.surrey.ac.uk Sat Jan 15 09:59:53 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Sat, 15 Jan 2000 09:59:53 +0000 (GMT) Subject: NS installation and configuration question In-Reply-To: Message-ID: On Fri, 14 Jan 2000, Haobo Yu wrote: > On FreeBSD tcl/tk are installed under different directories under > /usr/local/{include,lib} so ns will not mess up versions. It's unclear > what's your case, but you can always check tcl version by typing 'set > tcl_version' inside ns. Note ns does not need tk. ...even though it seems to be built using tk (grep tk ~ns/Makefile) and being able to issue tk commands in the interpreter would be Very Useful. assorted questions: 1) nam now handles gzipped files via zlib; very handy for saving diskspace. What chance xgraph could do the same for large datasets? 2) What's the status of http://www-mash.cs.berkeley.edu/doc-ns-2/ ? It doesn't appear to be linked to anywhere from the ns pages. cheers, L. PGP From L.Wood@eim.surrey.ac.uk Sat Jan 15 11:47:29 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Sat, 15 Jan 2000 11:47:29 +0000 (GMT) Subject: Mailing List Server Date In-Reply-To: <200001142057.MAA41971@destroyer.CS.Berkeley.EDU> Message-ID: On Fri, 14 Jan 2000, Brian K. Shiratsuki wrote: > Hypermail now produces correct dates for the January 2000 ns-users archives. I see mash's copy of hypermail has been upgraded to 2b28, and that's been run on January 2000. Thanks, Brian. The December 1999 archive index created with hypermail 2b3 is particularly incomplete, and certainly needs regenerating to list more than the 158 messages it lists. There are 237 messages rendered to HTML pages in the month, but the index stops at 8 December, even though the last line on the page says the archive was generated on 31 December... http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9912/ Can you regenerate the December 1999 archive from an archive mailbox using same, please? That should bring 'missing' messages in this month to light. (reading warnings on http://www.hypermail.org/y2k.html here, which might account for the 8/31 Dec discrepancy above.) I did some spot counts on other archive pages, and the results are worrying; the hypermail 1.02 index generator appears to be lossy too. For example: http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9911/ November 1999 (the last with hypermail 1.02) has 456 messages archived (0000.html - 0455.html). There are 458 lines of links on the index page according to grep -c href (each message title on a separate line), so if you subtract six lines to account for the context non-message links at top and bottom (more than one link a line), four messages remain unlisted. http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/9910/ October 1999 has 419 messages stored as individual webpages (0000.html - 0418.html), while grep -c \ 9910index.html gives 418 - but subtract the five non-message bullet points at top and bottom, and you're down to 413; five messages apparently not listed in the index. One of them is 0158.html. Another is 0342.html. The missing messages I located are both followups; I suspect that hypermail 1.02's threading code is unreliable, and that full regeneration of the archive would be prudent. (I suppose that, in grand IETF tradition, I should also ask for gzipped mailboxes to be made available from the ftp server so that the truly paranoid can generate their own indices.) cheers, L. is hoping that full regeneration fixes the two-digit years on: http://www-mash.CS.Berkeley.EDU/dist/archive/ns-users/ PGP From L.Wood@eim.surrey.ac.uk Sat Jan 15 12:18:18 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Sat, 15 Jan 2000 12:18:18 +0000 (GMT) Subject: searching mash for ns information Message-ID: http://www.ee.surrey.ac.uk/Personal/L.Wood/ns/ http://www.ee.surrey.ac.uk/Personal/L.Wood/ns/mash-search/ The mash htdig service has been running without allowing access to the first ten search results (header and footer, with links to subsequent results pages, are not generated.) So, I've hacked up these forms to allow access to more than the first ten results. I'd suggest putting a page selection menu in the existing htdig service as a stopgap. cheers, L. PGP From Martin Koyabe Sat Jan 15 06:09:08 2000 From: Martin Koyabe (Martin Koyabe) Date: Sat, 15 Jan 2000 06:09:08 +0000 (GMT) Subject: Code Error Message-ID: <200001150609.GAA24477@churchward.erg.abdn.ac.uk> Hi all, I am trying to compile some code in ns-2.1b6-current 'snapshot' and I get the following error at this line: ---code line--- # I am trying to set address for sender_addr set sender_addr [$sender set addr_] ---error message--- can't read "addr_": no such variable while executing "subst $[subst $var]" (procedure "_o84" line 5) (Object next line 5) invoked from within "_o84 next addr_" ("eval" body line 1) invoked from within "eval $self next $args" (procedure "_o84" line 18) (Agent set line 18) invoked from within "$mftp1 set addr_" (file "mftp_demo.tcl" line 87) ------ Any suggestions of possible cause of error !! -- Martin From haoboy@isi.edu Sat Jan 15 21:18:00 2000 From: haoboy@isi.edu (Haobo Yu) Date: Sat, 15 Jan 2000 13:18:00 -0800 (PST) Subject: Code Error In-Reply-To: <200001150609.GAA24477@churchward.erg.abdn.ac.uk> Message-ID: I had the impression that it's changed to dst_addr_ (as a result of the 32bit addressing). - Haobo On Sat, 15 Jan 2000, Martin Koyabe wrote: > Hi all, > > I am trying to compile some code in ns-2.1b6-current 'snapshot' and I get the > following error at this line: > > ---code line--- > > # I am trying to set address for sender_addr > > set sender_addr [$sender set addr_] > > > ---error message--- > can't read "addr_": no such variable > while executing > "subst $[subst $var]" > (procedure "_o84" line 5) > (Object next line 5) > invoked from within > "_o84 next addr_" > ("eval" body line 1) > invoked from within > "eval $self next $args" > (procedure "_o84" line 18) > (Agent set line 18) > invoked from within > "$mftp1 set addr_" > (file "mftp_demo.tcl" line 87) > ------ > > Any suggestions of possible cause of error !! > > -- Martin > > > From Martin Koyabe Sat Jan 15 06:50:06 2000 From: Martin Koyabe (Martin Koyabe) Date: Sat, 15 Jan 2000 06:50:06 +0000 (GMT) Subject: Code Error Message-ID: <200001150650.GAA24524@churchward.erg.abdn.ac.uk> Haobo, Not quite !! The 32bit addressing problem i.e. using dst_addr_ instead of dst_ is OK !! But take for example mftp_demo.tcl code from Chris's site http://www.cs.vu.nl/~chris/Thesis/mftp/mftp_readme.html and try to compile the code in ns-2.1b6 !! You might understand the error I am talking about !! -- Martin > > I had the impression that it's changed to dst_addr_ (as a result of the > 32bit addressing). > > - Haobo > > On Sat, 15 Jan 2000, Martin Koyabe wrote: > > > Hi all, > > > > I am trying to compile some code in ns-2.1b6-current 'snapshot' and I get the > > following error at this line: > > > > ---code line--- > > > > # I am trying to set address for sender_addr > > > > set sender_addr [$sender set addr_] > > > > > > ---error message--- > > can't read "addr_": no such variable > > while executing > > "subst $[subst $var]" > > (procedure "_o84" line 5) > > (Object next line 5) > > invoked from within > > "_o84 next addr_" > > ("eval" body line 1) > > invoked from within > > "eval $self next $args" > > (procedure "_o84" line 18) > > (Agent set line 18) > > invoked from within > > "$mftp1 set addr_" > > (file "mftp_demo.tcl" line 87) > > ------ > > > > Any suggestions of possible cause of error !! > > > > -- Martin > > > > > > > --------------------------------------------------------------------- Martin L.W.D Koyabe e-mail: koyabe@erg.abdn.ac.uk Electronics Research Group Fraser Noble Building King's College, AB24 3UE (Off) +44-01224-272813 Aberdeen, UK (Mob) +44-07881-610825 From haoboy@isi.edu Sat Jan 15 22:36:31 2000 From: haoboy@isi.edu (Haobo Yu) Date: Sat, 15 Jan 2000 14:36:31 -0800 (PST) Subject: Code Error In-Reply-To: <200001150650.GAA24524@churchward.erg.abdn.ac.uk> Message-ID: well, I have no idea about mftp, but if you look at agent.cc, you'll see that there're agnet_addr_ and dst_addr_. So if it's not dst_addr_ then it must be agent_addr_, which is the address of the agent itself. At a first glance I think mftp inherits these from Agent, so I guess that's what you want. - Haobo On Sat, 15 Jan 2000, Martin Koyabe wrote: > Haobo, > > Not quite !! > > The 32bit addressing problem i.e. using dst_addr_ instead of dst_ is OK !! But > take for example mftp_demo.tcl code from Chris's site > > http://www.cs.vu.nl/~chris/Thesis/mftp/mftp_readme.html > > and try to compile the code in ns-2.1b6 !! You might understand the error I am > talking about !! > > > -- Martin > > > > > I had the impression that it's changed to dst_addr_ (as a result of the > > 32bit addressing). > > > > - Haobo > > > > On Sat, 15 Jan 2000, Martin Koyabe wrote: > > > > > Hi all, > > > > > > I am trying to compile some code in ns-2.1b6-current 'snapshot' and I get > the > > > following error at this line: > > > > > > ---code line--- > > > > > > # I am trying to set address for sender_addr > > > > > > set sender_addr [$sender set addr_] > > > > > > > > > ---error message--- > > > can't read "addr_": no such variable > > > while executing > > > "subst $[subst $var]" > > > (procedure "_o84" line 5) > > > (Object next line 5) > > > invoked from within > > > "_o84 next addr_" > > > ("eval" body line 1) > > > invoked from within > > > "eval $self next $args" > > > (procedure "_o84" line 18) > > > (Agent set line 18) > > > invoked from within > > > "$mftp1 set addr_" > > > (file "mftp_demo.tcl" line 87) > > > ------ > > > > > > Any suggestions of possible cause of error !! > > > > > > -- Martin > > > > > > > > > > > > > --------------------------------------------------------------------- > Martin L.W.D Koyabe e-mail: koyabe@erg.abdn.ac.uk > Electronics Research Group > Fraser Noble Building > King's College, AB24 3UE (Off) +44-01224-272813 > Aberdeen, UK (Mob) +44-07881-610825 > > > From guru0109@palgong.knu.ac.kr Sun Jan 16 03:07:29 2000 From: guru0109@palgong.knu.ac.kr (=?EUC-KR?B?w9a/tbz2?=) Date: Sun, 16 Jan 2000 12:07:29 +0900 Subject: Addressing Structure in NS 2.1.b6? Message-ID: <004001bf5fce$d35f40c0$d50ee69b@knu.ac.kr> Hi. I use ns-allinone-2.1b5 and have problem about expanding node address format. recently i heard 2.1b6 use default 32 bit addressing is it true? i search change history but can't find any information about that. and if it is, what is default address format? how many bits for node id, port, multicast... etc . etc..? Thx in advance! ______________________________________________________ 弥康荐 (ChoiYoungSoo) e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 TeNet Lab : http://tenet.knu.ac.kr/ From engp9264@nus.edu.sg Sun Jan 16 04:41:48 2000 From: engp9264@nus.edu.sg (engp9264@nus.edu.sg) Date: Sun, 16 Jan 2000 12:41:48 +0800 Subject: open command Message-ID: <38814C0C.431C4744@nus.edu.sg> Is there limit about the files number of open command in Tcl? Thanks From Martin Koyabe Sat Jan 15 20:25:08 2000 From: Martin Koyabe (Martin Koyabe) Date: Sat, 15 Jan 2000 20:25:08 +0000 (GMT) Subject: Compilation error Message-ID: <200001152025.UAA25240@churchward.erg.abdn.ac.uk> Hi all, I am compiling my source code and keep getting this error: ---error--- Koyabe @ churchward 705 > !ns ns mftp_2.tcl Creating 2 Nodes ..... Creating Links ..... Transmitting 2000000 byte file with group-size 16 simulation is running... ns: _o112 start: can't read "dst_": no such variable while executing "expr $dst_" (procedure "_o112" line 4) (Agent/MFTP/Snd start line 4) invoked from within "_o112 start" ---end--- Can someone tell me what that error means ? -- Martin e-mail: koyabe@erg.abdn.ac.uk From Martin Koyabe Sun Jan 16 03:26:09 2000 From: Martin Koyabe (Martin Koyabe) Date: Sun, 16 Jan 2000 03:26:09 +0000 (GMT) Subject: Compilation error Message-ID: <200001160326.DAA00531@churchward.erg.abdn.ac.uk> Hi Developers, Just forwarding this over !! Could it be a bug in ns-2.1b6 or both mftp_snd/rcv.tcl files need appending !! I am compiling my source code and keep getting this error: ---error--- Koyabe @ churchward 705 > !ns ns mftp_2.tcl Creating 2 Nodes ..... Creating Links ..... Transmitting 2000000 byte file with group-size 16 simulation is running... ns: _o112 start: can't read "dst_": no such variable while executing "expr $dst_" (procedure "_o112" line 4) (Agent/MFTP/Snd start line 4) invoked from within "_o112 start" ---end--- Can someone tell me what that error means ? -- Martin e-mail: koyabe@erg.abdn.ac.uk From Martin Koyabe Sun Jan 16 03:42:35 2000 From: Martin Koyabe (Martin Koyabe) Date: Sun, 16 Jan 2000 03:42:35 +0000 (GMT) Subject: IP_BROADCAST warning message !! Message-ID: <200001160342.DAA00796@churchward.erg.abdn.ac.uk> Hi, Referring to: http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9908/0052.html was there a solution eliminating the recurrent 'IP_BROADCAST' warning message when compiling ns-2.1b6 ? I guess 'gcc' might not like the use of redefinition in solaris OS !!! Could some guide me on this !! -- Martin e-mail: koyabe@erg.abdn.ac.uk From c.christian.joensson@telia.com Sun Jan 16 19:09:10 2000 From: c.christian.joensson@telia.com (=?iso-8859-1?Q?Christian_J=F6nsson?=) Date: Sun, 16 Jan 2000 20:09:10 +0100 Subject: IP_BROADCAST warning message !! References: <200001160342.DAA00796@churchward.erg.abdn.ac.uk> Message-ID: <001201bf6055$2c06e800$59a7b5d4@chj> *If* I recall things correcty, it's a matter of moving the definition of IP_BROADCAST out of some *.cc file into some .h file and protect it using #ifndef. I think I submitted a suggestion of this to Nader Salehi in mid December 1999. I can check this up tomorrow at work, Cheers, /ChJ ----- Original Message ----- From: "Martin Koyabe" To: Cc: Sent: Sunday, January 16, 2000 4:42 AM Subject: IP_BROADCAST warning message !! > Hi, > > Referring to: > > http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9908/0052.html > > was there a solution eliminating the recurrent 'IP_BROADCAST' warning message > when compiling ns-2.1b6 ? I guess 'gcc' might not like the use of redefinition > in solaris OS !!! > > Could some guide me on this !! > > -- Martin > > e-mail: koyabe@erg.abdn.ac.uk > > From sitarama@cis.ohio-state.edu Mon Jan 17 01:33:07 2000 From: sitarama@cis.ohio-state.edu (vilayannur r sitaraman) Date: Sun, 16 Jan 2000 20:33:07 -0500 (EST) Subject: NS for optical networks? Message-ID: <200001170133.UAA21111@alpha.cis.ohio-state.edu> Hi, I am totally new to NS. Can NS be effectively used to simulate optical networks or should i be using a more appropriate simulation software(freely available) for this? Thanks a lot sitaraman From Mike Frumin Mon Jan 17 05:42:58 2000 From: Mike Frumin (Mike Frumin) Date: Sun, 16 Jan 2000 21:42:58 -0800 (PST) Subject: looking into packet headrers, please Message-ID: hi. I'm trying to modify the loss monitor to watch different multicast streams, but have a problem. in order to do this, it must look into the header o figure out the destination of the packet. I am having an EXTREMELY hard time figuring out how to do this, as I cant even find in the code how a packet is dropped onto the network using a CBR attached to a UDP agent. can someone please give me a hint? thanks, mike From Martin Koyabe Sun Jan 16 19:01:31 2000 From: Martin Koyabe (Martin Koyabe) Date: Sun, 16 Jan 2000 19:01:31 +0000 (GMT) Subject: Bug in ns-2.1b6 TCL files Message-ID: <200001161901.TAA05264@churchward.erg.abdn.ac.uk> Hi all, Just found that ~/tcl/mcast/mftp_rcv.tcl and ~/tcl/mcast/mftp_snd.tcl need to be updated as follows in order to compile well with 'dst_addr_' change !! replace 'dst_' with 'dst_addr_' in mftp_rcv.tcl [lines 44,46,48,57,59] also replace 'dst_' with 'dst_addr_' in mftp_snd.tcl [lines 65,67,70] Hope these changes will be updated in the next ns release !! -- Martin From forda@ite.mh.se Mon Jan 17 11:51:37 2000 From: forda@ite.mh.se (Daniel Forsgren) Date: Mon, 17 Jan 2000 12:51:37 +0100 (MET) Subject: Emulation problems Message-ID: Hello, I'm trying to use the NS emulator (nse) to evaluate the performance of different adaptive transmission schemes in a UDP/IP-based transmitter/receiver application pair, where data is sent in "layers" on separate ports. The nse host (a P3-550/FreeBSD 3.4 machine) serves as a gateway between the transmitter net and the receiver net. Things work reasonably well (although I do get an occasional "scheduler going backwards") until after a few minutes of emulation when everything breaks down. I get constant "going backwards" messages, mixed by "TapAgent(): sendpkt: No buffer space available"... and the receiver experience massive packetloss and jitter values. Strangely enough, things often calm down again after a few more minutes, and suddenly there is enough buffer space again. How can I resolve this problem? Can the buffer space be increased in some way? Thanks in advance, /Daniel F. From mark_wibble@hotmail.com Mon Jan 17 15:01:45 2000 From: mark_wibble@hotmail.com (Mark Butler) Date: Mon, 17 Jan 2000 07:01:45 PST Subject: Data handling in SRM Message-ID: <20000117150145.82864.qmail@hotmail.com> I am trying to create an agent that allows me to send actual data in SRM packets. I have so far created an agent called SRMD which is inherited from SRMAgent. This basically just handles the extra data in packets. I have also created an application SRMDApp to allow connection endpoints at nodes. This calls the sendmsg function of SRMD to send a packet containing data. The problem is, the packets appear in the nam trace as being sent out, but they are then dropped at the first nodes they get at, and prunes are sent to the sender. I have set the dst_ in the TCL script to be a common group. Does anyone have experience of doing this and can help ? Cheers, Mark ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From sher@dcs.warwick.ac.uk Mon Jan 17 15:20:53 2000 From: sher@dcs.warwick.ac.uk (Alex Poylisher) Date: Mon, 17 Jan 2000 15:20:53 +0000 Subject: a newbie question Message-ID: <200001171521.PAA00563@molly.dyndns.org> I'm looking into using ns to simulate network-aware mobile programs. Hence, the following rather generic question: is it possible to schedule topology modifications (e.g., introducing or removing links and nodes), and introduction/removal of agents and agent-to-node associations at points of time *after* the start of the simulation? -- Alex Poylisher PhD student, Computer Science, University of Warwick sher{at}dcs{dot}warwick{dot}ac{dot}uk From Kwang-Mien.Chan.GP@icn.siemens.de Mon Jan 17 15:57:49 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Mon, 17 Jan 2000 16:57:49 +0100 Subject: Snoop Message-ID: Hi there, i noticed that network simulator can simulate the SNOOP protocol as it has the snoop.cc file. Does anyone know how i can activate the SNOOP protocol in network simulator ? thanks. rgds, kwang mien From youngbae@cs.tamu.edu Mon Jan 17 17:14:33 2000 From: youngbae@cs.tamu.edu (Youngbae - Ko) Date: Mon, 17 Jan 2000 11:14:33 -0600 (CST) Subject: TORA with ns-2.1b6 ? Message-ID: Hello, ns users I wonder, one of ad hoc routing protocols (TORA) ported from CMU works fine with the current version of ns (ns-2.1b6) or not. I thought it would be, but at my preliminary trial I have been failed with the below error message: invalid command name "Agent/rtProto/TORA" while executing "Agent/rtProto/TORA instproc init args { puts "DOWN HERE" $self next $args } " (file "tcl/mobility/tora.tcl" line 21) I hope someone can help me to run TORA on the ns-2.1b6. Thanks, - Youngbae Ko Dept. of Computer Science Texas A&M University /* ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ */ From rsliu@csie.nctu.edu.tw Mon Jan 17 17:14:35 2000 From: rsliu@csie.nctu.edu.tw (Ren-Shiou Liu) Date: Tue, 18 Jan 2000 01:14:35 +0800 Subject: RSVP for NS-2.1b5 Message-ID: Hi, Dose anyone successfully run RSVP for NS-2.1b5 on Redhat Linux? Athough the compilation is OK. I got a segmentation fault error message. Can someone help me solve this problem? Thanks in advance! -- Jason Liu, Parallel Distributed Processing Lab. Dept. of Computer Science & Information Engineering National Chiao-Tung University, Hsinchu, Taiwan E-Mail: rsliu@csie.nctu.edu.tw From guille@iies.es Mon Jan 17 17:44:34 2000 From: guille@iies.es (Guillermo Rodriguez Garcia) Date: Mon, 17 Jan 2000 18:44:34 +0100 Subject: a newbie question Message-ID: <3.0.3.32.20000117184434.00d66a30@mailhost.iies.es> At 15:20 17/01/00 +0000, you wrote: > >I'm looking into using ns to simulate network-aware mobile programs. Hence, the >following rather generic question: is it possible to schedule topology modifications At the University of Malaga (Spain) we have developed a set of extensions to add support in ns for active networking, which is very related to the topic you mention above; the extensions are not finished yet, although they should be soon. A paper has been accepted at the Internet Workshop 2000 which will be held in Taiwan. We didn't consider dynamic topology changes, but our work might help you. Tell me if you are interested. Regards, G. -- Guillermo Rodriguez Garcia From lfoore@tantivy.com Mon Jan 17 18:29:46 2000 From: lfoore@tantivy.com (Larry Foore) Date: Mon, 17 Jan 2000 13:29:46 -0500 Subject: NS Simulation integration Message-ID: Hello, I am looking for some help with NS. I would like to use NS to generate traffic for a current custom simulation. The configuration will look something like the following: ____ ____ ____ / \ / \ / \ /TCP/IP\________/ OUR \________/TCP/IP\ \Client/ SLIP \ SIM / SLIP \SERVER/ \____/ \____/ \____/ NS NS Our simulator is a custom buit piece of code here that uses actual proprietary data link protocols and delivers TCP/IP packets between two wire stubs, as follows: (Note that this graphic is not the representation of an NS model.) ____ _____ ____ / \ /END \ / \ / WIRE \________/2 END \________/ WIRE \ \ STUB / STREAM \DLL SIM/ STREAM \ STUB / \____/ \_____/ \____/ The code in OUR SIM is to all be compiled externally so the actual code implementation can be used. This way, if the modifications of any code occurs, it can be recompiled and tested directly within the NS simulation. I have been trying to implement this using NS. It looks as if I will have to create a new node type that implements the network. This looks like a huge task in and of itself because I am not familiar with the internals of NS or C++ code. Basically, I need a node that can deliver packets to our simulator. Also, there is a timing issue. Is it possible to sync the sim clock with the clock executing the external code? Is it possible to dynamically change how much simulation time the external code consumes? Can the code for our simulation be sync'ed with NS time without altering the code (too much) ? I would appreciate any response. Thanks, Larry Foore Tantivy Communications, Inc. phone: (321) 956-8846 ext. 336 From s.a.janssen@student.utwente.nl Mon Jan 17 19:12:40 2000 From: s.a.janssen@student.utwente.nl (Sjoerd Janssen) Date: Mon, 17 Jan 2000 20:12:40 +0100 Subject: loss of original packet type info when tracing multiple protocols. Message-ID: <388369A8.4140B3CB@student.utwente.nl> Hi, - How does one retain the original information in the common header - the packet type being the most important- when using nodes consisting of multiple protocol layers and tracing for some of these layers is required? Consider for example a node containing a MAC layer and a TCP agent.. The ns/cmu trace objects generate their traces by looking at the packet type field in the common header. For a MAC packet, the type therefore needs to be changed from PT_TCP to PT_MAC in the sending MAC object (I know this is being done in mac-802_11.cc when a Packet is passed to the channel). On the receiving MAC side, the packet type should be changed back to PT_TCP but I don't see how this information can be retained. I also don't see how / if this is being done by the MAC object in mac-802_11.cc I guess there are ways to avoid it, e.g. packing the original (TCP/IP) packet into the data field of the Packet transmitted by MAC (and unpacking at the MAC receiver) or using an additional header to store the information contained in the original common header, but I'd really like to know how the people from CMU did this in mac-802_11 Regards, Sjoerd Janssen. From van8ive@hotmail.com Mon Jan 17 19:35:15 2000 From: van8ive@hotmail.com (Heather Reynolds) Date: Mon, 17 Jan 2000 19:35:15 GMT Subject: NS versus OPNET Modeler Message-ID: <20000117193515.7424.qmail@hotmail.com> I am negotiating a purchase of the OPNET Modeler software and a friend reccommended NS. Any old or current OPNET users? Who could give me feedback? Thanks, Heather ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From kkumar@isi.edu Mon Jan 17 19:43:17 2000 From: kkumar@isi.edu (Satish Kumar) Date: Mon, 17 Jan 2000 11:43:17 -0800 Subject: loss of original packet type info when tracing multiple protocols. In-Reply-To: Your message of Mon, 17 Jan 2000 20:12:40 +0100. <388369A8.4140B3CB@student.utwente.nl> Message-ID: <200001171943.LAA23325@gamma.isi.edu> Only packets generated by the mac-802.11 protocol such as RTS, CTS and ACK packets have PT_MAC as the packet type in the common header (if you look at mac-802_11.cc, you'll notice the packet type is set to PT_MAC only in sendRTS, sendCTS and sendACK methods). I dont think the packet type is changed to PT_MAC for packets generated by higher layer protocols. Satish > Hi, > > - How does one retain the original information in the common header - > the packet type being the most important- when using nodes consisting of > multiple protocol layers and tracing for some of these layers is > required? > > Consider for example a node containing a MAC layer and a TCP agent.. The > ns/cmu trace objects generate their traces by looking at the packet type > field in the common header. For a MAC packet, the type therefore needs > to be changed from PT_TCP to PT_MAC in the sending MAC object (I know > this is being done in mac-802_11.cc when a Packet is passed to the > channel). On the receiving MAC side, the packet type should be changed > back to PT_TCP but I don't see how this information can be retained. I > also don't see how / if this is being done by the MAC object in > mac-802_11.cc > > I guess there are ways to avoid it, e.g. packing the original (TCP/IP) > packet into the data field of the Packet transmitted by MAC (and > unpacking at the MAC receiver) or using an additional header to store > the information contained in the original common header, but I'd really > like to know how the people from CMU did this in mac-802_11 > > Regards, Sjoerd Janssen. > > > > > From s.a.janssen@student.utwente.nl Mon Jan 17 19:47:37 2000 From: s.a.janssen@student.utwente.nl (Sjoerd Janssen) Date: Mon, 17 Jan 2000 20:47:37 +0100 Subject: loss of original packet type info when tracing multiple protocols. References: <200001171943.LAA23325@gamma.isi.edu> Message-ID: <388371D9.3FF6F625@student.utwente.nl> Hi, thanks for your fast reply. You're right. But what do you think would be the best solution in case I do wish to use the ns/cmu trace objects to trace Packets at mulitple protocol layers? regards, Sjoerd Janssen. Satish Kumar wrote: > Only packets generated by the mac-802.11 protocol such as RTS, CTS and ACK > packets have PT_MAC as the packet type in the common header (if you look at > mac-802_11.cc, you'll notice the packet type is set to PT_MAC only in sendRTS, > sendCTS and sendACK methods). I dont think the packet type is changed to > PT_MAC for packets generated by higher layer protocols. > > Satish > > > Hi, > > > > - How does one retain the original information in the common header - > > the packet type being the most important- when using nodes consisting of > > multiple protocol layers and tracing for some of these layers is > > required? > > > > Consider for example a node containing a MAC layer and a TCP agent.. The > > ns/cmu trace objects generate their traces by looking at the packet type > > field in the common header. For a MAC packet, the type therefore needs > > to be changed from PT_TCP to PT_MAC in the sending MAC object (I know > > this is being done in mac-802_11.cc when a Packet is passed to the > > channel). On the receiving MAC side, the packet type should be changed > > back to PT_TCP but I don't see how this information can be retained. I > > also don't see how / if this is being done by the MAC object in > > mac-802_11.cc > > > > I guess there are ways to avoid it, e.g. packing the original (TCP/IP) > > packet into the data field of the Packet transmitted by MAC (and > > unpacking at the MAC receiver) or using an additional header to store > > the information contained in the original common header, but I'd really > > like to know how the people from CMU did this in mac-802_11 > > > > Regards, Sjoerd Janssen. > > > > > > > > > > From jscott@GEYSERNETWORKS.COM Mon Jan 17 22:07:21 2000 From: jscott@GEYSERNETWORKS.COM (James Scott) Date: Mon, 17 Jan 2000 14:07:21 -0800 Subject: NS versus OPNET Modeler Message-ID: <5C893CF12691D311B4E60050DA250700011CA7@QUEST> hi heather, > I am negotiating a purchase of the OPNET Modeler software > and a friend reccommended NS. Any old or current OPNET > users? Who could give me feedback? i have used both opnet and ns. the question of which is the best tool would depend on the problem you're trying to solve and the skills of the person who is doing the work. the process model OpNET provides is a powerful tool, providing sort of a 4GL for building simulation models of specific nodal processes. additionally, the opnet i/o is top notch. and, if your work involves modelling ATM nodes, opnet is clearly the way to go. also, mil3 has recently provided a VHDL i/f for OpNET, which would be quite handy for cosimulation applications. i've found NS to be superior in the area of source traffic generation. while OpNET does provide a flexible mechanism for generating various types of source traffic, NS has a larger number of source traffic generators, in addition to contributed models for higher layer protocols (HTTP, etc). however, since OpNET provides an open API for source traffic generation, there is nothing prohibiting you from using NS generators to stimulate an OpNET model, depending on the level of sophistication possessed by the developer doing the work. in the past, i've written pareto models to stimulate OpNET process models. had i known of NS at the time, i could have saved some time by leveraging their pareto model . in the area of routing, NS provides static and session-based routing using dijkstra, as well as dynamic routing using a distributed bellman-ford. by comparison, the OpNET tools provides just static routing. so, in summary, if you are modelling an IP network and interesting in things like queueing behavior and the interaction of higher-layer protocols on network performance, go with NS. you can use NS to model specific behaviors of proprietary h/w, but its perhaps not for the faint-of-heart. there are probably other areas where NS would be preferred, such as characterization of IP multicast and/or wireless IP, but i haven't had experience with these areas so will reserve comment. the above is just one man's opinion. both tools are QUITE useful, depending on the task at hand (as well as the size of your budget, and skills of your development crew). good luck! cheers, james From xwang@ctr.columbia.edu Mon Jan 17 23:35:32 2000 From: xwang@ctr.columbia.edu (Xin Wang) Date: Mon, 17 Jan 2000 18:35:32 -0500 (EST) Subject: NS versus OPNET Modeler In-Reply-To: <5C893CF12691D311B4E60050DA250700011CA7@QUEST> Message-ID: On Mon, 17 Jan 2000, James Scott wrote: > hi heather, > > > I am negotiating a purchase of the OPNET Modeler software > > and a friend reccommended NS. Any old or current OPNET > > users? Who could give me feedback? > > i have used both opnet and ns. the question of which is > the best tool would depend on the problem you're trying > to solve and the skills of the person who is doing the > work. > > the process model OpNET provides is a powerful tool, > providing sort of a 4GL for building simulation models > of specific nodal processes. additionally, the opnet i/o > is top notch. and, if your work involves modelling ATM > nodes, opnet is clearly the way to go. also, mil3 has > recently provided a VHDL i/f for OpNET, which would be > quite handy for cosimulation applications. > > i've found NS to be superior in the area of source traffic > generation. while OpNET does provide a flexible mechanism > for generating various types of source traffic, NS has a > larger number of source traffic generators, in addition to > contributed models for higher layer protocols (HTTP, etc). > however, since OpNET provides an open API for source traffic > generation, there is nothing prohibiting you from using NS > generators to stimulate an OpNET model, depending on the > level of sophistication possessed by the developer doing > the work. in the past, i've written pareto models to > stimulate OpNET process models. had i known of NS at the > time, i could have saved some time by leveraging their > pareto model . > > in the area of routing, NS provides static and session-based > routing using dijkstra, as well as dynamic routing using a > distributed bellman-ford. by comparison, the OpNET tools > provides just static routing. Opnet also provides RIP and OSPF routing, as well multicast routing recently. It is a more powerful simulation tool, the code is well organized and the tool is good for demo. But it has a bigger learning curve as trade-off. > so, in summary, if you are modelling an IP network > and interesting in things like queueing behavior and > the interaction of higher-layer protocols on network > performance, go with NS. you can use NS to model > specific behaviors of proprietary h/w, but its perhaps > not for the faint-of-heart. > > there are probably other areas where NS would be preferred, > such as characterization of IP multicast and/or wireless IP, > but i haven't had experience with these areas so will reserve > comment. the above is just one man's opinion. both tools > are QUITE useful, depending on the task at hand (as well as > the size of your budget, and skills of your development crew). > > good luck! > > cheers, > > james > From L.Wood@eim.surrey.ac.uk Mon Jan 17 23:53:21 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 17 Jan 2000 23:53:21 +0000 (GMT) Subject: NS versus OPNET Modeler In-Reply-To: Message-ID: On Mon, 17 Jan 2000, Xin Wang wrote: > Opnet also provides RIP and OSPF routing, as well multicast routing > recently. The multicast protocol functionality is not part of the standard package, and must be purchased separately. (I don't think it's available as part of their educational scheme.) See: http://www.mil3.com/products/library/specialized.html and note on: http://www.mil3.com/products/library/ Still, it claims to be an implementation of PIM-SM. (they're a bit keen on stressing that 'Study the benefits of using multicast versus unicast data transfer technologies'.) > It is a more powerful simulation tool, the code is > well organized and the tool is good for demo. > But it has a bigger learning curve as trade-off. and Opnet 6.0's interface is completely different to those of previous versions (grey browser-like look too); the satellite stuff has been hived off and now you must generate orbital models in STK and import them... Whole new learning curve. As a generalisation, Opnet's strengths lie lower in the network stack, but it's been working its way up. ns has been doing the reverse, fleshing out its structure downwards. The differences in engineering design/research focus emphasis should be noted, too. L. mind you, we've never had problems with our ns licence server. big ns plus. PGP From yaxu@isi.edu Tue Jan 18 00:56:15 2000 From: yaxu@isi.edu (Ya Xu) Date: Mon, 17 Jan 2000 16:56:15 -0800 (PST) Subject: TORA with ns-2.1b6 ? In-Reply-To: Message-ID: On Mon, 17 Jan 2000, Youngbae - Ko wrote: > Hello, ns users > > I wonder, one of ad hoc routing protocols (TORA) ported from CMU > works fine with the current version of ns (ns-2.1b6) or not. yes > > I thought it would be, but at my preliminary trial > I have been failed with the below error message: > > invalid command name "Agent/rtProto/TORA" > while executing > "Agent/rtProto/TORA instproc init args { > puts "DOWN HERE" > $self next $args > } " > (file "tcl/mobility/tora.tcl" line 21) > > try test-suite-wireless-lan-tora.tcl under tcl/test From kkumar@isi.edu Tue Jan 18 02:42:51 2000 From: kkumar@isi.edu (Satish Kumar) Date: Mon, 17 Jan 2000 18:42:51 -0800 Subject: loss of original packet type info when tracing multiple protocols. In-Reply-To: Your message of Mon, 17 Jan 2000 20:47:37 +0100. <388371D9.3FF6F625@student.utwente.nl> Message-ID: <200001180242.SAA16040@gamma.isi.edu> If you set MacTrace, AgentTrace or RouterTrace ON in your tcl script, cmu-trace objects are appropriately inserted (look at the add-target MobileNode method in tcl/lib/ns-mobilenode.tcl). You might want to look at ns Notes and Documentation or any existing ad hoc protocol implementation such as DSDV for ideas. Satish > Hi, thanks for your fast reply. You're right. But what do you think would be the > best solution in case I do wish to use the ns/cmu trace objects to trace Packets > at mulitple protocol layers? > > regards, Sjoerd Janssen. > > > > Satish Kumar wrote: > > > Only packets generated by the mac-802.11 protocol such as RTS, CTS and ACK > > packets have PT_MAC as the packet type in the common header (if you look at > > mac-802_11.cc, you'll notice the packet type is set to PT_MAC only in sendRTS, > > sendCTS and sendACK methods). I dont think the packet type is changed to > > PT_MAC for packets generated by higher layer protocols. > > > > Satish > > > > > Hi, > > > > > > - How does one retain the original information in the common header - > > > the packet type being the most important- when using nodes consisting of > > > multiple protocol layers and tracing for some of these layers is > > > required? > > > > > > Consider for example a node containing a MAC layer and a TCP agent.. The > > > ns/cmu trace objects generate their traces by looking at the packet type > > > field in the common header. For a MAC packet, the type therefore needs > > > to be changed from PT_TCP to PT_MAC in the sending MAC object (I know > > > this is being done in mac-802_11.cc when a Packet is passed to the > > > channel). On the receiving MAC side, the packet type should be changed > > > back to PT_TCP but I don't see how this information can be retained. I > > > also don't see how / if this is being done by the MAC object in > > > mac-802_11.cc > > > > > > I guess there are ways to avoid it, e.g. packing the original (TCP/IP) > > > packet into the data field of the Packet transmitted by MAC (and > > > unpacking at the MAC receiver) or using an additional header to store > > > the information contained in the original common header, but I'd really > > > like to know how the people from CMU did this in mac-802_11 > > > > > > Regards, Sjoerd Janssen. > > > > > > > > > > > > > > > > From Kwang-Mien.Chan.GP@icn.siemens.de Tue Jan 18 10:16:34 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Tue, 18 Jan 2000 11:16:34 +0100 Subject: New Node API Message-ID: hi there, i downloaded the daily snapshot for NS and installed it successfully. i also downloaded the special snapshot (having debugged version of CMU's wireless module and other wireless extensions) and installed it. when i tried to run the simple-wireless.tcl example in Marc Greis磗 tutorial, the following error message appeared. unable to dispatch method node-config ... when i checked the /ns/tcp/lib/ns-lib.tcl file, there are no definition of node-config. i also realised that if i do not install CMU codes, i would have the definition of node-config after installing the daily snapshot. so, i tried not installing the CMU codes. however, installation of the daily snapshot would not be successful without the CMU codes. does anyone have any idea how i could get the New Node API working ? thanks. rgds, kwang mien From George.Khoury@era.ericsson.se Tue Jan 18 10:34:47 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Tue, 18 Jan 2000 11:34:47 +0100 Subject: Segmentation in TCP Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE00E@esealnt190> Dears ns-usres, I wonder if there is any TCP agent that can break down a large packet size into smaller packet size. It would be more efficient if the TCP agent didn't add the hole header and waited for the ack for each packet. I mean in that case it would like if I decided a smaller packet size. Any comment would be appreciated. George From ana_delgado77@hotmail.com Tue Jan 18 15:41:08 2000 From: ana_delgado77@hotmail.com (Ana Delgado) Date: Tue, 18 Jan 2000 15:41:08 GMT Subject: Multicast for Mobility Protocols Message-ID: <20000118154108.42647.qmail@hotmail.com> Hi everyone, I would like to know if an implementation of Multicast for Mobility Protocols exists in ns. Thanks in advance, Ana ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From yaxu@isi.edu Tue Jan 18 15:59:33 2000 From: yaxu@isi.edu (Ya Xu) Date: Tue, 18 Jan 2000 07:59:33 -0800 (PST) Subject: New Node API In-Reply-To: Message-ID: The debugged version of CMU wireless module is out of date. Please do not use it w/ current ns snapshot. The current snapshot should support node-config -Ya On Tue, 18 Jan 2000, Chan Kwang Mien wrote: > > hi there, > > i downloaded the daily snapshot for NS and installed it successfully. i > also downloaded the special snapshot (having debugged version of CMU's > wireless module and other wireless extensions) and installed it. when i > tried to run the simple-wireless.tcl example in Marc Greis磗 tutorial, the > following error message appeared. > > unable to dispatch method node-config > ... > > when i checked the /ns/tcp/lib/ns-lib.tcl file, there are no definition > of node-config. i also realised that if i do not install CMU codes, i would > have the definition of node-config after installing the daily snapshot. so, > i tried not installing the CMU codes. however, installation of the daily > snapshot would not be successful without the CMU codes. > > does anyone have any idea how i could get the New Node API working ? > thanks. > > rgds, > kwang mien > From guru0109@palgong.knu.ac.kr Tue Jan 18 17:11:28 2000 From: guru0109@palgong.knu.ac.kr (=?EUC-KR?B?w9a/tbz2?=) Date: Wed, 19 Jan 2000 02:11:28 +0900 Subject: Half Duplex or Full Duplex? Message-ID: <00fc01bf61d7$0ebd1980$d50ee69b@knu.ac.kr> Hi. I have question about link. I find 2 link type duplexlink and simplex link I wonder duplex link in NS is fulll duplex? or half duplex? And... additionally, I heard that router send some message(maybe ICMP ?) to senders when sender transmitt UDP(overlimit link speed) is it true? and if it is, NS support that figure? ______________________________________________________ 弥康荐 (ChoiYoungSoo) e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 TeNet Lab : http://tenet.knu.ac.kr/ From L.Wood@eim.surrey.ac.uk Tue Jan 18 17:23:35 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Tue, 18 Jan 2000 17:23:35 +0000 (GMT) Subject: Half Duplex or Full Duplex? In-Reply-To: <00fc01bf61d7$0ebd1980$d50ee69b@knu.ac.kr> Message-ID: On Wed, 19 Jan 2000, [EUC-KR] 弥康荐 wrote: > I find 2 link type duplexlink and simplex link > I wonder duplex link in NS is fulll duplex? or half duplex? full duplex; a duplex link is actually a wrapper that creates two similar simplex links, one in either direction, with their heads in different nodes. If you want half-duplex, set up a LAN. > And... additionally, > I heard that router send some message(maybe ICMP ?) to senders when sender > transmitt UDP(overlimit link speed) > is it true? and if it is, NS support that figure? ICMP source quench? It's rarely used in the Real World AFAIK. L. http://palgong.kyungpook.ac.kr/%7Eguru0109/work.html has some handy local-to-Korea mirrors of things. PGP From haoboy@isi.edu Tue Jan 18 17:43:06 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 18 Jan 2000 09:43:06 -0800 (PST) Subject: Segmentation in TCP In-Reply-To: <5F05C89FB2F8D211B6430008C7919127020EE00E@esealnt190> Message-ID: FullTcp does segmentation and reassembly, if that is what you meant. - Haobo On Tue, 18 Jan 2000, George Khoury (ERA) wrote: > Dears ns-usres, > > I wonder if there is any TCP agent that can break down a large packet size into smaller packet size. It would be more efficient if the TCP agent didn't add the hole header and waited for the ack for each packet. I mean in that case it would like if I decided a smaller packet size. > > Any comment would be appreciated. > > George > > From puginifabio@tiscalinet.it Tue Jan 18 17:50:41 2000 From: puginifabio@tiscalinet.it (Fabio Pugini) Date: Tue, 18 Jan 2000 18:50:41 +0100 Subject: Installation help Message-ID: <000701bf61dc$8ee47720$98850b3e@default> This is a multi-part message in MIME format. ------=_NextPart_000_0004_01BF61E4.EB859060 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Please help me! I've just downloaded the all-in-one version of ns in order to install it = on my home computer but I can't succeed in doing it! The install file and all the directories ( cweb,sgb,xgraph......) are in = /root/Desktop/NS/ns-allinone2.1b5. Which are the right commands to install ns in /root/Desktop/NS ? ( I've read all the instructions but it really doesn't work !!!) =20 I'll be very grateful to anyone will answer. Fabio Pugini ------=_NextPart_000_0004_01BF61E4.EB859060 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Please help me!
I've just = downloaded the=20 all-in-one version of ns in order to install it on my home computer but = I can't=20 succeed in doing it!
The install file and all the directories (=20 cweb,sgb,xgraph......) are in /root/Desktop/NS/ns-allinone2.1b5.
Which are the right commands to install ns in = /root/Desktop/NS=20 ?
( I've read all the instructions but it  really = doesn't=20 work !!!)
 
I'll be very = grateful to=20 anyone will answer.
Fabio Pugini
 
 
------=_NextPart_000_0004_01BF61E4.EB859060-- From salehi@isi.edu Tue Jan 18 21:01:42 2000 From: salehi@isi.edu (Nader Salehi) Date: Tue, 18 Jan 2000 13:01:42 -0800 (PST) Subject: Compilation error In-Reply-To: <200001152025.UAA25240@churchward.erg.abdn.ac.uk> References: <200001152025.UAA25240@churchward.erg.abdn.ac.uk> Message-ID: <200001182101.NAA11691@joy.isi.edu> Martin, Take a look at agent.cc and agent.h and you'll see that dst_ is decmposed into daddr_ and dport_. Nader >>> "Martin" == Martin Koyabe writes: Martin> Hi all, I am compiling my source code and keep getting this Martin> error: Martin> ---error--- Martin> Koyabe @ churchward 705 > !ns ns mftp_2.tcl Creating 2 Nodes Martin> ..... Creating Links ..... Transmitting 2000000 byte file Martin> with group-size 16 simulation is running... ns: _o112 start: Martin> can't read "dst_": no such variable while executing "expr Martin> $dst_" (procedure "_o112" line 4) (Agent/MFTP/Snd start line Martin> 4) invoked from within "_o112 start" Martin> ---end--- Martin> Can someone tell me what that error means ? Martin> -- Martin Martin> e-mail: koyabe@erg.abdn.ac.uk From klan@usc.edu Tue Jan 18 21:11:13 2000 From: klan@usc.edu (Kun-Chan Lan) Date: Tue, 18 Jan 2000 13:11:13 -0800 Subject: Installation help References: <000701bf61dc$8ee47720$98850b3e@default> Message-ID: <3884D6F1.62FAF3E5@isi.edu> --------------3AA22FF0A1060F804D322BAC Content-Type: text/plain; charset=big5 Content-Transfer-Encoding: 7bit Fabio Pugini wrote: > Please help me!I've just downloaded the all-in-one version of ns in > order to install it on my home computer but I can't succeed in doing > it!The install file and all the directories ( cweb,sgb,xgraph......) > are in /root/Desktop/NS/ns-allinone2.1b5.Which are the right commands > to install ns in /root/Desktop/NS ?( I've read all the instructions > but it really doesn't work !!!) I'll be very grateful to anyone will > answer.Fabio Pugini run ./install from /root/Desktop/NS/ns-allinone2.1b5/ Kun-chan Lan --------------3AA22FF0A1060F804D322BAC Content-Type: text/html; charset=big5 Content-Transfer-Encoding: 7bit Fabio Pugini wrote:
 Please help me!I've just downloaded the all-in-one version of ns in order to install it on my home computer but I can't succeed in doing it!The install file and all the directories ( cweb,sgb,xgraph......) are in /root/Desktop/NS/ns-allinone2.1b5.Which are the right commands to install ns in /root/Desktop/NS ?( I've read all the instructions but it  really doesn't work !!!) I'll be very grateful to anyone will answer.Fabio Pugini


run ./install   from /root/Desktop/NS/ns-allinone2.1b5/

Kun-chan Lan --------------3AA22FF0A1060F804D322BAC-- From Martin Koyabe Tue Jan 18 06:40:32 2000 From: Martin Koyabe (Martin Koyabe) Date: Tue, 18 Jan 2000 06:40:32 +0000 (GMT) Subject: Compilation error Message-ID: <200001180640.GAA07655@churchward.erg.abdn.ac.uk> Nader, You'll be suprised what I foung to be the problem !! I found that ~/tcl/mcast/mftp_rcv.tcl and ~/tcl/mcast/mftp_snd.tcl need to be updated as follows in order to compile well with 'dst_addr_' change !! replace 'dst_' with 'dst_addr_' in mftp_rcv.tcl [lines 44,46,48,57,59] also replace 'dst_' with 'dst_addr_' in mftp_snd.tcl [lines 65,67,70] Hope these changes will be updated in the next ns release !! -- Martin > > Martin, > > Take a look at agent.cc and agent.h and you'll see that dst_ is > decmposed into daddr_ and dport_. > > Nader > > > >>> "Martin" == Martin Koyabe writes: > > Martin> Hi all, I am compiling my source code and keep getting this > Martin> error: > > Martin> ---error--- > > Martin> Koyabe @ churchward 705 > !ns ns mftp_2.tcl Creating 2 Nodes > Martin> ..... Creating Links ..... Transmitting 2000000 byte file > Martin> with group-size 16 simulation is running... ns: _o112 start: > Martin> can't read "dst_": no such variable while executing "expr > Martin> $dst_" (procedure "_o112" line 4) (Agent/MFTP/Snd start line > Martin> 4) invoked from within "_o112 start" > > Martin> ---end--- > > Martin> Can someone tell me what that error means ? > > Martin> -- Martin > > Martin> e-mail: koyabe@erg.abdn.ac.uk > From salehi@isi.edu Tue Jan 18 21:19:08 2000 From: salehi@isi.edu (Nader Salehi) Date: Tue, 18 Jan 2000 13:19:08 -0800 (PST) Subject: Compilation error In-Reply-To: <200001180640.GAA07655@churchward.erg.abdn.ac.uk> References: <200001180640.GAA07655@churchward.erg.abdn.ac.uk> Message-ID: <200001182119.NAA11707@joy.isi.edu> Martin, Thanks for pointing out the problem. Since the mftp source does not seem to be part of our validation test scripts, these errors continued to exist. Nader >>> "Martin" == Martin Koyabe writes: Martin> Nader, You'll be suprised what I foung to be the problem !! Martin> I found that ~/tcl/mcast/mftp_rcv.tcl and Martin> ~/tcl/mcast/mftp_snd.tcl need to be updated as follows in Martin> order to compile well with 'dst_addr_' change !! Martin> replace 'dst_' with 'dst_addr_' in mftp_rcv.tcl [lines Martin> 44,46,48,57,59] also replace 'dst_' with 'dst_addr_' in Martin> mftp_snd.tcl [lines 65,67,70] Martin> Hope these changes will be updated in the next ns release !! Martin> -- Martin >> Martin, >> >> Take a look at agent.cc and agent.h and you'll see that dst_ is >> decmposed into daddr_ and dport_. >> >> Nader >> >> >> >>> "Martin" == Martin Koyabe writes: >> Martin> Hi all, I am compiling my source code and keep getting this Martin> error: >> Martin> ---error--- >> Martin> Koyabe @ churchward 705 > !ns ns mftp_2.tcl Creating 2 Nodes Martin> ..... Creating Links ..... Transmitting 2000000 byte file Martin> with group-size 16 simulation is running... ns: _o112 start: Martin> can't read "dst_": no such variable while executing "expr Martin> $dst_" (procedure "_o112" line 4) (Agent/MFTP/Snd start line Martin> 4) invoked from within "_o112 start" >> Martin> ---end--- >> Martin> Can someone tell me what that error means ? >> Martin> -- Martin >> Martin> e-mail: koyabe@erg.abdn.ac.uk >> From litmaath@fnal.gov Tue Jan 18 22:29:42 2000 From: litmaath@fnal.gov (Maarten Litmaath) Date: Tue, 18 Jan 2000 16:29:42 -0600 Subject: 512 parallel FTP sessions very slow? Message-ID: <3884E956.90835C76@fnal.gov> Dear NS experts, we are in the process of evaluating NS for simulating a 512x512 switch (we know that NS does not support switches yet, so that we would have to write some code ourselves; will that change, by the way?). To get an idea of the speed of NS we tried the following little script: it runs 512 parallel FTP sessions. We found that on a Sparc-20 it took 1.5 hours for 1 ms of simulation!!! Are we doing something wrong? Thanks, Maarten (Fermilab CD/CMS dept.) #---------------------------------------------------------------------- set ns [new Simulator] $ns set-address-format expanded set maxnode 512 for {set i 0} {$i < $maxnode} {incr i} { set node($i) [$ns node] set k [expr $i + $maxnode] set node($k) [$ns node] $ns duplex-link $node($i) $node($k) 640Mb 0.1ms DropTail $ns queue-limit $node($i) $node($k) 10 set tcp [$ns create-connection TCP $node($i) TCPSink $node($k) 0] set ftp [$tcp attach-source FTP] $ns at 0.000 "$ftp start" $ns at 0.001 "$ftp stop" } $ns at 0.002 "$ns halt" $ns run #---------------------------------------------------------------------- From guille@iies.es Tue Jan 18 23:00:27 2000 From: guille@iies.es (Guillermo Rodriguez Garcia) Date: Wed, 19 Jan 2000 00:00:27 +0100 Subject: 512 parallel FTP sessions very slow? In-Reply-To: <3884E956.90835C76@fnal.gov> Message-ID: <3.0.3.32.20000119000027.0082a100@mailhost.iies.es> At 16:29 18/01/00 -0600, you wrote: >Dear NS experts, >we are in the process of evaluating NS for simulating a 512x512 switch >(we know that NS does not support switches yet, so that we would have to >write some code ourselves; will that change, by the way?). >To get an idea of the speed of NS we tried the following little script: >it runs 512 parallel FTP sessions. >We found that on a Sparc-20 it took 1.5 hours for 1 ms of simulation!!! >Are we doing something wrong? A good fraction of this time is due to the topology generation - nodes and links, not related to the duration of the simulation itself. Regards, G. -- Guillermo Rodriguez Garcia From yiyung@mmlab.snu.ac.kr Wed Jan 19 02:10:47 2000 From: yiyung@mmlab.snu.ac.kr (Yung Yi) Date: Wed, 19 Jan 2000 11:10:47 +0900 Subject: tcp trace Message-ID: <002101bf6222$671b0f60$e3d50198@apan.snu.ac.kr> Hi. I want to know the tcp behavior after getting tcp trace. I mean cwnd, ssthreshold and something like that. Could someone please give me the information about that? I think I can use bin/xxx.pl, too. However, I don't know how to use it. When I see the files in test/ directory, it stores all trace(all.tr) and uses getrc and raw2xg. Getrc and raw2xg is enough to get all the statistics of tcp's behavior. I want to know how to use the files in bin/ directory. Thanks in advance. ------------------------------------------------------ Yung Yi Multimedia & Computer Communication Lab. Dept. of Computer Engineering, Seoul National Univ. Tel : +82-2-876-7170 Fax : +82-2-876-7171 Email : yiyung@mmlab.snu.ac.kr URL : http://mmlab.snu.ac.kr/~yiyung ------------------------------------------------------ From salehi@isi.edu Wed Jan 19 02:30:23 2000 From: salehi@isi.edu (Nader Salehi) Date: Tue, 18 Jan 2000 18:30:23 -0800 (PST) Subject: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released Message-ID: <200001190230.SAA22658@joy.isi.edu> What is NS? ---------- NS is a discrete event simulator targeted at networking research. NS provides substantial support for simulation of TCP, routing, and multicast protocols. NS began as a variant of the REAL network simulator in 1989 and has evolved substantially over the past few years. The NS development effort is now an ongoing collaboration with the VINT project. What is Nam? ---------- Nam is a Tcl/TK based animation tool for viewing network simulation traces and real world packet traces. It supports topology layout, packet level animation, and various data inspection tools. Nam began at LBL. It has evolved substantially over the past few years. The nam development effort is now an ongoing collaboration with the VINT project. WINDOWS USES: Currently Nam does NOT work with Tcl/TK 8.0.5 on Windows 95/98/NT. If you want to build nam there, please use the Tcl/TK 8.0p2. In order to build a statically linked nam on the Windows platforms, you need to download a patched version of Tcl/TK 8.0p2 from http://mash.cs.berkeley.edu/dist/, and follow the instructions in their respective makefile.vc. What is NS-ALLINONE? ----------- Ns-allinone is a package which contains required components and some of optional components used in running NS. The package contains an "install" script to automatically configure, compile and install these components. After downloading, run the install script. If you haven't installed ns before and want to quickly try ns out, ns-allinone may be easier than getting all the pieces by hand. Currently the package contains: Tcl release 8.0.4 (required component) Tk release 8.0.4 (required component) Otcl release 1.0a5 (required component) TclCL release 1.0b9 (required component) Ns release 2.1b6 (required component) Nam release 1.0a8 (optional component) Zlib version 1.1.3 (optional component, required should Nam be used) Xgraph version 12 (optional component) CWeb version 3.4g (optional component) SGB version 1.0 (?) (optional component, builds sgblib for all UNIX type platforms) Gt-itm gt-itm & sgb2ns 1.1(optional component) What is New in NS 2.1b5? ------------------ Some of the new features in this version include: - Modified the real-time scheduler (used in nse) to reset the real time clock to zero when executing "$ns run", which is synchronized with the simulator's virtual clock. - Cleared the SACK scoreboard when reseting a SACK TCP connection - Implemented delay binding for TCP TraceVar instance variables - Added RFC 793 TCP implementation with some add-ons for educational purposes. - Changed the nodes' internal addressing structure to 32-bit. - Added new trace format to wireless trace. - Added configurable parameters for AIMD increase and decrease constants in TCP. - Replaced regular classifier with the hash classifier. - Re-implemented wiredLAN(802.3) by porting cmu's 802.3 module - Added a fix to generate nam traces correctly for packets dropped by error models. - Added mobileIP support. - Changed autoconf handling of perl - Many bug fixes. - And many more! What is new in NAM 1.0a7? -------------------- New features in this version include: - Added node size scale up/down buttons in the zoom bar. - Replaced the edit view menu entry with a toolbar button - Fixed the node (inner) label color. - Added support for large flow id and color ids. - Added mechanism for visualization of packet flow in wireless simulation. - Re-instated options '-j' and '-r' to control the startup time and animation rate. - Allowed users to provide initialization file for every source file. NS 2.1b6 can be downloaded from and NAM 1.0a8 can be downloaded from and NS-ALLINONE 2.1b6 can be downloaded from and Thanks, Nader Salehi USC/ISI From lauhow@cs.curtin.edu.au Wed Jan 19 03:10:11 2000 From: lauhow@cs.curtin.edu.au (William Lau) Date: Wed, 19 Jan 2000 11:10:11 +0800 Subject: exhausted UIDs Message-ID: <020001bf622a$b34427c0$ffbcfea9@cs.curtin.edu.au> Hi all, ns complains that i've exhausted UID space after running my simulation for a few hours. It seems as if ns only allows a finite number of events to be generated in one simulation. I would be grateful if someone could tell me what i am doing wrong with my code or how i can circumvent this. Any pointers/info appreciated. cheers William From guru0109@palgong.knu.ac.kr Wed Jan 19 06:45:22 2000 From: guru0109@palgong.knu.ac.kr (=?euc-kr?B?w9a/tbz2?=) Date: Wed, 19 Jan 2000 15:45:22 +0900 Subject: Installation Problem ns-allinone2.1b5 on Solaris. Bug FIx dosn't work! Message-ID: <001301bf6248$c24a9ba0$d50ee69b@knu.ac.kr> Hi. here is my error message -----snip---- checking for a BSD compatible install... ./install-sh -c creating ./config.status creating Makefile rm -f embedded-tcl.cc ./tcl2c++ et_tcl ../tclbox/lib/tcl8.0/init.tcl > embedded-tcl.cc *** Error code 139 make: Fatal error: Command failed for target `embedded-tcl.cc' tclcl-1.0b8 make failed! Exiting ... tenet3 /usr/home/guru0109/ns-allinone-2.1b5 (10) % -----snip----- I have some problem when install tcl, tk, otcl, but --enable-gcc solve those. But when i try to "make" tclcl, error!!! Of course I search Bug Fix... So i add "setenv LD_LIBRARY_PATH /usr/local/lib........." and i verify it works. In bug fix, 2 solutions about embedded.cc 1. Install gcc or 2. patch... I try to both, but doesn't work... How should i do now? ______________________________________________________ 弥康荐 (ChoiYoungSoo) e-m@ail : mailto:guru0109@palgong.knu.ac.kr icq : 43998959 myh0me : http://palgong.knu.ac.kr/~guru0109 TeNet Lab : http://tenet.knu.ac.kr/ From Kwang-Mien.Chan.GP@icn.siemens.de Wed Jan 19 09:55:29 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Wed, 19 Jan 2000 10:55:29 +0100 Subject: proxy arp Message-ID: hi there, does anyone know how i can send proxy ARP in NS ? thanks. rgds, kwang mien From George.Khoury@era.ericsson.se Wed Jan 19 12:00:13 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Wed, 19 Jan 2000 13:00:13 +0100 Subject: Pareto!! Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE011@esealnt190> Dears ns-users, I have a small question regarding Application/Traffic/Pareto: I Wonder if the packet size in constant as for Exponential traffic!? Is there any easy way to make the packet size pareto distributed too!!? Thank in advance George From L.Wood@eim.surrey.ac.uk Wed Jan 19 12:45:06 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 19 Jan 2000 12:45:06 +0000 (GMT) Subject: Compilation error In-Reply-To: <200001182119.NAA11707@joy.isi.edu> Message-ID: On Tue, 18 Jan 2000, Nader Salehi wrote: > Thanks for pointing out the problem. Since the mftp source does not > seem to be part of our validation test scripts, these errors continued > to exist. Oh, this is So Wrong. Example scripts supplied with ns should never drop into the tcl debugger. If they do, there's Obviously a Problem That Needs Fixing. Setting up a test harness that runs through every example script supplied with ns once every couple of weeks or so, and alerts developers to error messages and debugger output (which shouldn't be happening) would be a very sensible thing to do. It's not the behavioural catch-all of nightly validation checks, and it's not intended to be. But it would be a good way of finding and catching problems early before they impact the userbase; the example scripts are part of the ns package, they do use code that validation doesn't catch, and they should be tested. (I've also noticed that one reaction to clarify something that's not clear to an ns user is to add comments in the relevant validation test, rather than to an example script or the documentation.) L. PGP From Kwang-Mien.Chan.GP@icn.siemens.de Wed Jan 19 13:05:06 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Wed, 19 Jan 2000 14:05:06 +0100 Subject: snoop problem Message-ID: hi there, > i got the following error message after i run this > > ns snoop.tcl -e 32768 -eu byte -llSnoop -r > > (_o63 cmd line 1) > invoked from within > "_o63 cmd id" > invoked from within > "catch "$self cmd $args" ret" > (procedure "_o63" line 2) > (Split Object unknown line 2) > ..... > > i am using the daily snapshot (18/1/2000 ) for NS with the CMU's > wireless module and other wireless extensions installed. do u know how i > could solve this problem ? > > thank you. > > rgds, > kwang mien From ppieda@nortelnetworks.com Wed Jan 19 13:18:16 2000 From: ppieda@nortelnetworks.com (Peter Pieda) Date: Wed, 19 Jan 2000 07:18:16 -0600 Subject: What is Pareto distribution? Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF627F.A6AD61D4 Content-Type: text/plain Hello, Check out this Prof's Statistics web Page at Osaka University: http://www.sigmath.es.osaka-u.ac.jp/~aki/pdf/pdf.htm It gives the equation for the distribution and then gives a few example graphs. Peter > -----Original Message----- > From: HUANG PENG [SMTP:huangpeng_buaa@a-1.net.cn] > Sent: Sunday, January 19, 1997 8:00 PM > To: ns-users > Subject: What is Pareto distribution? > > Hi, everyone: > > Can you give me a description of Pareto distribution? > > Thank you in advance ------_=_NextPart_001_01BF627F.A6AD61D4 Content-Type: text/html Content-Transfer-Encoding: quoted-printable RE: What is Pareto distribution?

Hello,

Check out this = Prof's Statistics web Page at Osaka University:

http://www.sigmath.es.osaka-u.ac.jp/~aki/pdf/pdf.htm

It gives the = equation for the distribution and then gives a few example
graphs.

Peter


    -----Original Message-----
    From:   HUANG PENG = [SMTP:huangpeng_buaa@a-1.net.cn]
    Sent:   Sunday, January 19, 1997 8:00 PM
    To:     ns-users
    Subject:       = What is Pareto distribution?

    Hi, everyone:
     
     Can you give me a description = of Pareto distribution?

     Thank you in advance

------_=_NextPart_001_01BF627F.A6AD61D4-- From aries@kungsrv.ee.ncku.edu.tw Wed Jan 19 16:43:18 2000 From: aries@kungsrv.ee.ncku.edu.tw (Aries) Date: Thu, 20 Jan 2000 00:43:18 +0800 Subject: RSVP model Message-ID: <000801bf629c$4a0ab860$bf48748c@ee.ncku.edu.tw> This is a multi-part message in MIME format. ------=_NextPart_000_0005_01BF62DF.5804C580 Content-Type: text/plain; charset="big5" Content-Transfer-Encoding: quoted-printable Gentleman , We want to study about RSVP and use ns to build RSVP = model . But we lack knowledge=20 and experience of these topics. Would you send me some examples ? Thanks = for your help. = cheers, C.H.Ke Ultra Speed Networking and Multimedia Architecture Lab, ncku, Taiwan,ROC ------=_NextPart_000_0005_01BF62DF.5804C580 Content-Type: text/html; charset="big5" Content-Transfer-Encoding: quoted-printable
Gentleman ,
          &nbs= p;    =20 We want to study about RSVP and use ns to build RSVP model . But we lack = knowledge
and experience of these topics. Would you send me = some=20 examples ? Thanks for your help.
 
           &n= bsp;           &nb= sp;           &nbs= p;            = ;            =             &= nbsp;           &n= bsp;   cheers,
 
C.H.Ke
 
Ultra Speed Networking and Multimedia Architecture Lab,
ncku, Taiwan,ROC
 
------=_NextPart_000_0005_01BF62DF.5804C580-- From Tarik Alj Wed Jan 19 17:35:19 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 19 Jan 2000 12:35:19 -0500 (EST) Subject: RSVP model Message-ID: <200001191735.MAA27625@someware.INRS-Telecom.UQuebec.CA> http://www.teltec.dcu.ie/~murphys Sean Murphy has ported Marc Greiss's RSVP module to ns-2.1b5; more info can be obtained by searching the list archives. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > From: "Aries" > To: > Subject: RSVP model > Date: Thu, 20 Jan 2000 00:43:18 +0800 > MIME-Version: 1.0 > X-Priority: 3 > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2615.200 > > Gentleman , > We want to study about RSVP and use ns to build RSVP model . But we lack knowledge > and experience of these topics. Would you send me some examples ? Thanks for your help. > > cheers, > > C.H.Ke > > Ultra Speed Networking and Multimedia Architecture Lab, > ncku, Taiwan,ROC > Tarik From Tarik Alj Wed Jan 19 18:38:16 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 19 Jan 2000 13:38:16 -0500 (EST) Subject: rtp Message-ID: <200001191838.NAA01641@someware.INRS-Telecom.UQuebec.CA> Hi, it seems ns's page at http//www-mash.cs.berkeley.edu/ns is currently unavailable. So i hope the answer to my question is not in the list's archives... Is it possible to use RTP in a unicast scheme? If so, does anybody have example scripts or pointers? Thanks in advance, Tarik From haoboy@isi.edu Wed Jan 19 18:57:38 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 19 Jan 2000 10:57:38 -0800 (PST) Subject: Pareto!! In-Reply-To: <5F05C89FB2F8D211B6430008C7919127020EE011@esealnt190> Message-ID: Pareto traffic model, as well as exponential, models traffic by packet bursts and they are irrelevant to packet size. If you want model packet size, it should be straightforward to build a traffic model following the example of CBR; all you need to do is to change the packet size in each delivery (override TrafficGenerator::timeout(), or provide your own virtual send()). - Haobo On Wed, 19 Jan 2000, George Khoury (ERA) wrote: > Dears ns-users, > > > I have a small question regarding Application/Traffic/Pareto: > I Wonder if the packet size in constant as for Exponential traffic!? > Is there any easy way to make the packet size pareto distributed too!!? > > Thank in advance > > George > > From huiwenli@nortelnetworks.com Wed Jan 19 21:34:11 2000 From: huiwenli@nortelnetworks.com (Huiwen Li) Date: Wed, 19 Jan 2000 13:34:11 -0800 Subject: half-duplex port? Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF62C4.EC64ECD4 Content-Type: text/plain; charset="ISO-8859-1" Hi, Does ns implement half-duplex port or anyone have experience in implementing? I have the following topology n0 --- \ \ n2 --------n3 / / n1 --- n2 node has two ports, one connecting to n0 and the other to n1, the link connecting n0 and n2 and the link connecting n1 and n2 are half-duplex. So both ports at n2 are half-duplex. I know I should implement the link as a LAN, but n2 itself should be in each LAN, otherwise n0 and n1 will be connected as in the same LAN. I thought I could insert a node between n0 and n2 to act as a "virtual port", but the buffer setting would be very tricky. Does any one have an idea how to implement this? Thanks, Huiwen ------_=_NextPart_001_01BF62C4.EC64ECD4 Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable half-duplex port?

Hi,

Does ns implement half-duplex port or = anyone have experience in implementing? I have the following = topology

n0 ---
          \
           = ; \
           = ;   n2 --------n3
           = ; /
          /
n1 ---

n2 node has two ports, one connecting = to n0 and the other to n1, the link connecting n0 and n2 and the link = connecting n1 and n2 are half-duplex. So both ports at n2 are = half-duplex. I know I should implement the link as a LAN, but n2 itself = should be in each LAN, otherwise n0 and n1 will be connected as in the = same LAN. I thought I could insert a node between n0 and n2 to act as a = "virtual port", but the buffer setting would be very tricky. =

Does any one have an idea how to = implement this?

Thanks,
Huiwen


------_=_NextPart_001_01BF62C4.EC64ECD4-- From Tarik Alj Wed Jan 19 22:13:00 2000 From: Tarik Alj (Tarik Alj) Date: Wed, 19 Jan 2000 17:13:00 -0500 (EST) Subject: half-duplex port? Message-ID: <200001192213.RAA07944@someware.INRS-Telecom.UQuebec.CA> Huiwen, set lan1 [$ns make-lan "$n0 $n2" args] set lan2 [$ns make-lan "$n1 $n2" args] set the_link [$ns duplex-link $n2 $n3 args] should do it... use plain Mac, not Mac/802_3 > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > From: "Huiwen Li" > To: "'ns-users@mash.cs.berkeley.edu'" > Subject: half-duplex port? > Date: Wed, 19 Jan 2000 13:34:11 -0800 > MIME-Version: 1.0 > > Hi, > > Does ns implement half-duplex port or anyone have experience in > implementing? I have the following topology > > n0 --- > \ > \ > n2 --------n3 > / > / > n1 --- > > n2 node has two ports, one connecting to n0 and the other to n1, the link > connecting n0 and n2 and the link connecting n1 and n2 are half-duplex. So > both ports at n2 are half-duplex. I know I should implement the link as a > LAN, but n2 itself should be in each LAN, otherwise n0 and n1 will be > connected as in the same LAN. I thought I could insert a node between n0 and > n2 to act as a "virtual port", but the buffer setting would be very tricky. > > Does any one have an idea how to implement this? > > Thanks, > Huiwen > > Tarik From xu003@hotmail.com Wed Jan 19 23:28:14 2000 From: xu003@hotmail.com (Shugong Xu) Date: Wed, 19 Jan 2000 18:28:14 EST Subject: the validation test problem for ns2 Message-ID: <20000119232814.47896.qmail@hotmail.com> hi, there, I have installed successfully the ns2 snapshot of Dec 7, 1999. However, it outputs the following messages when I run the validation test suite: validate overall report: some tests failed: ./test-all-wireless-lan ./test-all-wireless-lan-newnode to re-run a specific test, cd tcl/test; ../../ns test-all-TEST-NAME /home/johnh/WORKING/VINT/test-ns-2/test_ns: validate I noticed that the snapshot after Dec 21 had fixed it from the following document: http://www.isi.edu/~johnh/WORK/VINT/VALIDATION/199912/ Is there anone have some patchs or something else to help me fix this problem? Thanks in advance. Shugong ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From yiyung@mmlab.snu.ac.kr Thu Jan 20 03:16:21 2000 From: yiyung@mmlab.snu.ac.kr (Yung Yi) Date: Thu, 20 Jan 2000 12:16:21 +0900 Subject: retransmission timeout Message-ID: <002101bf62f4$bc8d91c0$e3d50198@apan.snu.ac.kr> Hi. in traceVar function of tcp.cc, we can trace cwnd_,.... values. However, that code does not contain retransmission timeout value(t_rtxcur_). I want to trace out retransmission timeout value change of tcp connection. How can I do it? If you have information about this problem, Please send the e-mail to me. Thanks. ------------------------------------------------------ Yung Yi Multimedia & Computer Communication Lab. Dept. of Computer Engineering, Seoul National Univ. Tel : +82-2-876-7170 Fax : +82-2-876-7171 Email : yiyung@mmlab.snu.ac.kr URL : http://mmlab.snu.ac.kr/~yiyung ------------------------------------------------------ From engp9264@nus.edu.sg Thu Jan 20 07:28:25 2000 From: engp9264@nus.edu.sg (engp9264@nus.edu.sg) Date: Thu, 20 Jan 2000 15:28:25 +0800 Subject: retransmission timeout References: <002101bf62f4$bc8d91c0$e3d50198@apan.snu.ac.kr> Message-ID: <3886B919.8CC51820@nus.edu.sg> Yung Yi wrote: > Hi. > > in traceVar function of tcp.cc, we can trace cwnd_,.... values. > However, that code does not contain retransmission timeout value(t_rtxcur_). > I want to trace out retransmission timeout value change of tcp connection. > > How can I do it? > > If you have information about this problem, Please send the e-mail to me. > > Thanks. > ------------------------------------------------------ > Yung Yi > Multimedia & Computer Communication Lab. > Dept. of Computer Engineering, Seoul National Univ. > > Tel : +82-2-876-7170 > Fax : +82-2-876-7171 > Email : yiyung@mmlab.snu.ac.kr > URL : http://mmlab.snu.ac.kr/~yiyung > ------------------------------------------------------ You need to change the "double t_rtxcur_ " to "TracedDouble t_rtxcur_" in tcp.h at the first. Then you need to add some source code to record your variable in TcpAgent::traceVar fuction in the file of tcp.cc At last you need to recomplie the ns. I am not sure it can success. Just try and then tell me the results,ok? Xiaoyun From swtan@mmu.edu.my Thu Jan 20 16:21:28 2000 From: swtan@mmu.edu.my (Tan Su Wei) Date: Thu, 20 Jan 2000 16:21:28 +0000 Subject: Questions about ManualRtNode, hash classifier Message-ID: <38873608.51EF182D@mmu.edu.my> This is a multi-part message in MIME format. --------------E79BCF82100C2BBB593892A7 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Dear all, I'd tried to simulate a manual route packet forwarding using ManualRtNode {in ns-node.tcl}, I got some error when trying to run a simple simulation script. I can figure out what is wrong with the script and no sure what the error message means. Can some one please show me how to used ManualRtNode or the hashclassifier for setting up a pre comfigure path. Attach is my simulation script... and below is the error message generated.... Regards Tan Su Wei --------------E79BCF82100C2BBB593892A7 Content-Type: application/x-tcl; name="learnClassifier.tcl" Content-Transfer-Encoding: 7bit Content-Disposition: inline; filename="learnClassifier.tcl" #simulator instance set ns [new Simulator] $ns color 1 blue $ns color 2 red #Open file for nam trace file set nf [open out1.nam w] $ns namtrace-all $nf #Finish procedure proc finish {} { global ns nf $ns flush-trace #close trace file close $nf #execute nam exec nam out1.nam & exit 0 } #define topology # # r2--r5 # / \ # n1---r1 r4---n2 # \ / # \_r3_/ # # packet to be routed from: n1-r1-r2-r5-r4-n2 ########################## #Define nodes set n1 [new ManualRtNode] set n2 [new ManualRtNode] set r1 [new ManualRtNode] set r2 [new ManualRtNode] set r3 [new ManualRtNode] set r4 [new ManualRtNode] set r5 [new ManualRtNode] #Define links $ns duplex-link $n1 $r1 1Mb 10ms DropTail $ns duplex-link $r1 $r2 1Mb 10ms DropTail $ns duplex-link $r2 $r5 1Mb 10ms DropTail $ns duplex-link $r1 $r3 1Mb 10ms DropTail $ns duplex-link $r5 $r4 1Mb 10ms DropTail $ns duplex-link $r3 $r4 1Mb 10ms DropTail $ns duplex-link $r4 $n2 1Mb 10ms DropTail # manual set route $n1 add-route-to-adj-node $r1 $r1 add-route-to-adj-node $r2 $r2 add-route-to-adj-node $r5 $r5 add-route-to-adj-node $r4 $r4 add-route-to-adj-node $n2 #Define agent set udp1 [new Agent/UDP] $ns attach-agent $n1 $udp1 set cbr1 [new Application/Traffic/CBR] $cbr1 set packetSize_ 100 $cbr1 set rate_ 1Mb $cbr1 set fid_ 1 $cbr1 attach-agent $udp1 set udp2 [new Agent/UDP] $ns attach-agent $n1 $udp2 set cbr2 [new Application/Traffic/CBR] $cbr2 set packetSize_ 100 $cbr2 set rate_ 1Mb $cbr2 set fid_ 2 $cbr2 attach-agent $udp2 set sink1 [new Agent/LossMonitor] $ns attach-agent $n2 $sink1 $ns connect $udp1 $sink1 set sink2 [new Agent/LossMonitor] $ns attach-agent $n2 $sink2 $ns connect $udp2 $sink2 # event $ns at 0.05 "$cbr1 start" $ns at 0.10 "$cbr2 start" $ns at 5.0 "$cbr1 stop" $ns at 4.0 "$cbr2 stop" $ns at 5.05 "finish" $ns run --------------E79BCF82100C2BBB593892A7-- From Martin Koyabe Wed Jan 19 19:31:30 2000 From: Martin Koyabe (Martin Koyabe) Date: Wed, 19 Jan 2000 19:31:30 +0000 (GMT) Subject: No subject Message-ID: <200001191931.TAA10322@churchward.erg.abdn.ac.uk> Hi all, Has anyone tried simulating Multicast to big user groups (500, 1000, 10000) ? using tiers as the random network topology generator (a partfrom Chris Hanle's work)? I guess my swap space and RAM might pose some constrain !! -- Martin From Martin Koyabe Wed Jan 19 19:33:12 2000 From: Martin Koyabe (Martin Koyabe) Date: Wed, 19 Jan 2000 19:33:12 +0000 (GMT) Subject: Multicast Scaling Message-ID: <200001191933.TAA10328@churchward.erg.abdn.ac.uk> Hi all, Has anyone tried simulating Multicast to big user groups (500, 1000, 10000) ? using tiers as the random network topology generator (a partfrom Chris Hanle's work)? I guess my swap space and RAM might pose some constrain !! -- Martin From Kwang-Mien.Chan.GP@icn.siemens.de Thu Jan 20 11:46:48 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Thu, 20 Jan 2000 12:46:48 +0100 Subject: snoop problem Message-ID: hi there, > i seem to have problem with the result generated > by snoop.tcl. > > When i run snoop.tcl by using the 2 commands below, > the Y-value seems to be constant at 2Mb for 5 seconds. the bandwidth of > the link is 2 Mb . my mac is set to 802_3 since the Csma/Ca would > generate errors. > > ns snoop.tcl -e32768 -eubyte -stop 5 -llSnoop > ns snoop.tcl -e0.001 -eutime -stop 5 -llSnoop > > is the result generated correct ? > thanks. > > rgds, > mien > > > From savoric@ftmail.ee.TU-Berlin.DE Thu Jan 20 11:55:02 2000 From: savoric@ftmail.ee.TU-Berlin.DE (Michael Savoric) Date: Thu, 20 Jan 2000 12:55:02 +0100 Subject: No subject Message-ID: <00012013050900.08108@hyperion> Hi, I have written a tcl-function in which I want to execute the ns-function at. How can I use in the code segment of the at-function the local variables of my calling function? upvar -1 or upvar #1 doesn't work. And only using global variables prevents the reusing of my function. Thanks. Best regards, Michael Savoric From engp9264@nus.edu.sg Thu Jan 20 12:38:16 2000 From: engp9264@nus.edu.sg (Liu Xiaoyun) Date: Thu, 20 Jan 2000 20:38:16 +0800 Subject: Message-ID: <30A14FB41CC5D311854D00508B5EEF02011893BC@exs23.ex.nus.edu.sg> -----Original Message----- From: Michael Savoric To: ns-users@mash.CS.Berkeley.EDU Sent: 1/20/00 7:55 PM Hi, I have written a tcl-function in which I want to execute the ns-function at. How can I use in the code segment of the at-function the local variables of my calling function? upvar -1 or upvar #1 doesn't work. And only using global variables prevents the reusing of my function. Thanks. Best regards, Michael Savoric $self instvar yourvaluename From Tarik Alj Thu Jan 20 13:40:04 2000 From: Tarik Alj (Tarik Alj) Date: Thu, 20 Jan 2000 08:40:04 -0500 (EST) Subject: Questions about ManualRtNode, hash classifier Message-ID: <200001201340.IAA04482@someware.INRS-Telecom.UQuebec.CA> Tan, i think the probleme comes from: # manual set route $n1 add-route-to-adj-node $r1 $r1 add-route-to-adj-node $r2 $r2 add-route-to-adj-node $r5 $r5 add-route-to-adj-node $r4 $r4 add-route-to-adj-node $n2 the unknown flow error comes from the classifier being unable to map the flow from 0 to 256 with fid of 0 (1 stands for the bucket number) to a slot. Either you are not using these intsproc correctly or you've stepped across a bug? > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Thu, 20 Jan 2000 16:21:28 +0000 > From: Tan Su Wei > X-Accept-Language: en > MIME-Version: 1.0 > To: "ns-users@mash.CS.Berkeley.EDU" > Subject: Questions about ManualRtNode, hash classifier > > Dear all, > I'd tried to simulate a manual route packet forwarding using > ManualRtNode {in ns-node.tcl}, I got some error when trying to run a > simple simulation script. > I can figure out what is wrong with the script and no sure what the > error message means. Can some one please show me how to used > ManualRtNode or the hashclassifier for setting up a pre comfigure path. > > Attach is my simulation script... and below is the error message > generated.... > > ns: _o11 unknown-flow 0 256 0 1: > (_o11 cmd line 1) > invoked from within > "_o11 cmd unknown-flow 0 256 0 1" > invoked from within > "catch "$self cmd $args" ret" > (procedure "_o11" line 2) > (SplitObject unknown line 2) > invoked from within > "_o11 unknown-flow 0 256 0 1" > --------------------------> > > Regards > Tan Su Wei Tarik From George.Khoury@era.ericsson.se Thu Jan 20 13:41:04 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Thu, 20 Jan 2000 14:41:04 +0100 Subject: Pareto Shape Parameter Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE015@esealnt190> Dears ns-users, Could any body explain what is the shape parameter for pareto distribution? Any suggestions would be appreciated Thanks George From George.Khoury@era.ericsson.se Thu Jan 20 13:43:45 2000 From: George.Khoury@era.ericsson.se (George Khoury (ERA)) Date: Thu, 20 Jan 2000 14:43:45 +0100 Subject: Help Modified expoo.cc Message-ID: <5F05C89FB2F8D211B6430008C7919127020EE016@esealnt190> Dears ns-users, I trying to implement an exponential traffic with pareto packet size. I thought that I could copy the exponential traffic (expoo.cc) and just call the function pareto to generate size_, but when I run nam I see that the packet size is always 1000 (probably it comes from TCP/Agent). Could any body give some suggestion why that happens and how to solve the problem. Any suggestion will be appreciated Thanks in advance George From mark_wibble@hotmail.com Thu Jan 20 16:36:58 2000 From: mark_wibble@hotmail.com (Mark Butler) Date: Thu, 20 Jan 2000 08:36:58 PST Subject: Naming user defined packets Message-ID: <20000120163658.73064.qmail@hotmail.com> In the NAM simulations, TCP packets etc have names when clicked on. Is there a way to name a user defined packet as well ? Cheers, Mark ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From haoboy@isi.edu Thu Jan 20 17:49:43 2000 From: haoboy@isi.edu (Haobo Yu) Date: Thu, 20 Jan 2000 09:49:43 -0800 (PST) Subject: your mail In-Reply-To: <200001191931.TAA10322@churchward.erg.abdn.ac.uk> Message-ID: You can look at http://www-mash.CS.Berkeley.EDU/ns/ns-debugging.html for some memory tips. You can switch to session-level routing if you don't care details. Polly Huang has reached 50,000 nodes in about 200 groups using that. - Haobo On Wed, 19 Jan 2000, Martin Koyabe wrote: > > Hi all, > > Has anyone tried simulating Multicast to big user groups (500, 1000, 10000) ? > using tiers as the random network topology generator (a partfrom Chris Hanle's > work)? I guess my swap space and RAM might pose some constrain !! > > > -- Martin From mallman@grc.nasa.gov Thu Jan 20 19:24:10 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Thu, 20 Jan 2000 14:24:10 -0500 Subject: a though to help with debugging ns Message-ID: <200001201924.OAA03531@lombok-fi.lerc.nasa.gov> A long sad story, followed by a hopefully helpful suggestion... I have been running some fairly large simulations recently. Something like 500-1000 FTP sessions, which yield many more TCP flows. The simulations take ~20 minutes to run and produce many, many megabytes of trace files. I started running across "bugs". I started seeing connections that would start, but never finish. Or, connections that were supposed to start but never sent any packets. A bunch of things started jumping out after I distilled the trace file down to something I could look at. Some of these bugs were obviously with my distilling code. But, others turned out to be ns' fault (which I have posted to the list). I am sure it is not news to a lot of you that debugging a 20 minute simulation with thousands of connections is no fun task. My first thought was to try to come up with a small simulation of one TCP connection that would run in 2 seconds and then I could debug that with gdb or print statements or whatever. No such luck. And, I tried hard. After it was clear that I was going to have to debug the 20 minute simulation I went home for the day had a beer and decided to tackle it the next morning. Half asleep the next morning I thought gdb with watchpoints was my answer. Obviously I had not taken in enough caffeine yet. Watchpoints take forever on a small program, let alone a 20 minute simulation. I decided to keep thinking... So, I thought I'd just stick some print statements around (the first debugging technique I learned and still by far the most robust). Obviously that was not a well thought out idea either as I now got output dumps that were bigger than the trace files (many megabytes). Sifting through them was going to be painful at best. Next I remembered that I knew when the event was supposed to happen (or did happen -- depending on the particular bug I was chasing). So, I added a little if statement in front of my debugging print statements to only print stuff around the time when the error was happening. Still lots of garbage about connections that were working perfectly fine. Finally, it hit me... What I needed was a variable in the TCP class that I could turn on if I wanted the debugging. And, so that is what I did... So, I added a variable to the protected section of the TcpAgent class... int tcp_debug_; And, in the ns-default.tcl file I added a default value... Agent/TCP set tcp_debug_ false Finally, in TcpAgent::TcpAgent() I added... bind_bool ("tcp_debug_", &tcp_debug_); Then, all debugging type statements take a form like... if (tcp_debug_) printf ("slowdown()\n"); Now the debugging information can be turned on and off (assuming you're using callback routines, which I am) in the tcl code. Very nice. This generally helped point out the errors rather rapidly. Well, OK, it still took 20 minutes for each simulation run, but I had to make very few debugging runs after I introduced this machinery, since the debugging output was easy to read and only for the busted connection. I suggested the above to Sally, who thought it was a nice idea. However, while this took about 2 minutes to implement the above, what I am suggesting is that ns include these debugging variables all over the place (I have subsequently added a sink_debug_ to the TCP sink class). For what it's worth, I encourage the VINT team to add tons of such variables. Maybe not all in one pass, but as you work on various classes keep this in the back of your mind. I am guessing that would be a big help in debugging rare problems that only seem to crop up in large simulations. allman --- http://roland.grc.nasa.gov/~mallman/ From huiwenli@nortelnetworks.com Thu Jan 20 20:20:11 2000 From: huiwenli@nortelnetworks.com (Huiwen Li) Date: Thu, 20 Jan 2000 12:20:11 -0800 Subject: Questions about LAN Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF6383.C0810AF2 Content-Type: text/plain; charset="ISO-8859-1" Hi, I have two questions about LAN: 1. How can I access ifq_ for one of the nodes on LAN? I want to monitor the ifq_. 2. If I want to use MAC/802_3, must I define the bandwidth to be 10Mbps? What if I want a lan with similar features to MAC/802_3 with bandwidth as 100Mbps? Hope to get some answers as soon as possible. Thanks! Huiwen ------_=_NextPart_001_01BF6383.C0810AF2 Content-Type: text/html; charset="ISO-8859-1" Content-Transfer-Encoding: quoted-printable Questions about LAN

Hi,

I have two questions about = LAN:

1. How can I access ifq_ for one of = the nodes on LAN? I want to monitor the ifq_.

2. If I want to use MAC/802_3, must I = define the bandwidth to be 10Mbps? What if I want a lan with similar = features to MAC/802_3 with bandwidth as 100Mbps?

Hope to get some answers as soon as = possible. Thanks!

Huiwen

------_=_NextPart_001_01BF6383.C0810AF2-- From L.Wood@eim.surrey.ac.uk Thu Jan 20 20:41:16 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Thu, 20 Jan 2000 20:41:16 +0000 (GMT) Subject: a though to help with debugging ns In-Reply-To: <200001201924.OAA03531@lombok-fi.lerc.nasa.gov> Message-ID: On Thu, 20 Jan 2000, Mark Allman wrote: [..] > Finally, it hit me... What I needed was a variable in the TCP class > that I could turn on if I wanted the debugging. And, so that is > what I did... > > So, I added a variable to the protected section of the TcpAgent > class... > > int tcp_debug_; [..] > I suggested the above to Sally, who thought it was a nice idea. > However, while this took about 2 minutes to implement the above, > what I am suggesting is that ns include these debugging variables > all over the place (I have subsequently added a sink_debug_ to the > TCP sink class). Great idea, but... Okay, extremely stupid implementation question and nitpick here. Since the variables are class variables and have local scope, why not just be clear and consistent and call them *all* debug_ ? Agent/TCP set debug_ true has got to be different from Agent/TCPSink set debug_ true and at the OTcl level: set tcp666 [new Agent/TCP/Vegas/UpperDorking] [..] set debugging_my_bloody_tcp_instance_again [$tcp666 set debug_ true] [..] if { debugging_my_bloody_tcp_instance_again } { puts "this time for sure!" [..] } seems straightforward enough, even to me. The thought of scouring the ns code to find out what the specific debug variable you're supposed to be using around your printf statements is called, or getting it wrong and misinterpreting the output as a result, really doesn't appeal. (Reminds me of the occasional legacy problem resulting from not checking that the Tcl variable is bound to a C++ variable of the same name and CaSe_, in fact...) Can't inheritance take care of it, so we can have globally available debug_ functionality via a consistently-named instance variable? Just define and bind debug_ false in Agent and the other base classes and let everything inherit from those, and a lot should be done. L. I mean, all this overhyped object-oriented twaddle I keep hearing about has to be good for _something_. > For what it's worth, I encourage the VINT team to add tons of such > variables. Maybe not all in one pass, but as you work on various > classes keep this in the back of your mind. I am guessing that > would be a big help in debugging rare problems that only seem to > crop up in large simulations. > > allman > > > --- > http://roland.grc.nasa.gov/~mallman/ PGP From klan@usc.edu Thu Jan 20 21:08:37 2000 From: klan@usc.edu (Kun-chan Lan) Date: Thu, 20 Jan 2000 13:08:37 -0800 Subject: Naming user defined packets Message-ID: <000401bf638a$88a95a60$0100a8c0@trojan.usc.edu> you can change the value of parameter "-p" in nam trace file by modifing ~ns/trace.cc or ~ns/cmu-trace.cc (for wireless) kun-chan -----Original Message----- From: Mark Butler To: ns-users@mash.cs.berkeley.edu Date: Thursday, January 20, 2000 8:43 AM Subject: Naming user defined packets >In the NAM simulations, TCP packets etc have names when clicked on. Is there >a way to name a user defined packet as well ? > >Cheers, > >Mark >______________________________________________________ >Get Your Private, Free Email at http://www.hotmail.com > From Tarik Alj Thu Jan 20 22:18:11 2000 From: Tarik Alj (Tarik Alj) Date: Thu, 20 Jan 2000 17:18:11 -0500 (EST) Subject: Questions about LAN Message-ID: <200001202218.RAA17633@someware.INRS-Telecom.UQuebec.CA> When using make-lan, the value of bw you are passing as a parameter will be the value of the LAN's bw, no matter what mac you are using. ifq_ is stored in class LanIface in turn stored in lanIface_; so you can do something like: set nif [$lan set lanIface_($node)] set ifq [$nif set ifq_] if you want to use 802.3 with bw values other than 10Mbps you should check with the corresponding values of IEEE for backoffLimit, attemplimit, jamSize, slotTime and interFrameGap. They will change for 100Mbps and 1000Mbps. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > From: "Huiwen Li" > To: "'ns-users@mash.cs.berkeley.edu'" > Subject: Questions about LAN > Date: Thu, 20 Jan 2000 12:20:11 -0800 > MIME-Version: 1.0 > > Hi, > > I have two questions about LAN: > > 1. How can I access ifq_ for one of the nodes on LAN? I want to monitor the > ifq_. > > 2. If I want to use MAC/802_3, must I define the bandwidth to be 10Mbps? > What if I want a lan with similar features to MAC/802_3 with bandwidth as > 100Mbps? > > Hope to get some answers as soon as possible. Thanks! > > Huiwen > Tarik Alj INRS-Telecommunications Place Bonaventure 900 De La Gauchetierre Ouest Niveau C, Case Postale 644 Montreal, Qc, H5A 1C6 Canada From Fusun Inanc Thu Jan 20 22:51:38 2000 From: Fusun Inanc (Fusun Inanc) Date: Thu, 20 Jan 2000 17:51:38 -0500 (EST) Subject: ErrorModel rate_??? Message-ID: <200001202251.RAA05818@poptop.> Dear ns_users, I am using ErrorModel agent in my simulation. And I want to create a link which has a bit error rate of 1e-08. If I set the rate $loss_module1 set rate_ 1e-05 it drops packets. But if I set it to 1e-06 , 1e-07 and so on, It doesn't do anything. I run the simulation 50 sec. Is there a way to make the error rate more sensitive? Thanks in advance... Fusun Inanc From L.Wood@eim.surrey.ac.uk Thu Jan 20 23:15:37 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Thu, 20 Jan 2000 23:15:37 +0000 (GMT) Subject: ErrorModel rate_??? In-Reply-To: <200001202251.RAA05818@poptop.> Message-ID: On Thu, 20 Jan 2000, Fusun Inanc wrote: > I am using ErrorModel agent in my simulation. And I want to > create a link which has a bit error rate of 1e-08. > > If I set the rate > > $loss_module1 set rate_ 1e-05 > > it drops packets. If you don't specify the error rate as anything, it's packet rate (pkt), not bit rate. I think you need to translate your BER into an approximately equivalent byte error rate, and do $lossmodule unit byte first. (the end of nsDoc 12.1 says something about "either in packets of bits", which is extremely misleading, and imo needs fixing.) L. > But if I set it to 1e-06 , 1e-07 and so on, It doesn't do anything. > I run the simulation 50 sec. > > Is there a way to make the error rate more sensitive? > > Thanks in advance... > > Fusun Inanc PGP From engp9374@nus.edu.sg Fri Jan 21 05:07:13 2000 From: engp9374@nus.edu.sg (Ge Yu) Date: Fri, 21 Jan 2000 13:07:13 +0800 Subject: RSVP simulation in 2.1b6 Message-ID: <30A14FB41CC5D311854D00508B5EEF0201ED54E8@exs23.ex.nus.edu.sg> hi, friends, When I want to update RSVP code(2.1b5) to 2.1b6, I met with some problems, I have changed the address-variables to confirm to b6, the "make" is successful, but when I run rsvp_conf.tcl, the errors occur as follows: _o3: unable to dispatch method get-nodetype while executing "[Simulator instance] get-nodetype" (procedure "_o10" line 6) (Node init line 6) invoked from within "_o10 init {}" (Class create line 1) invoked from within "Node create _o10 {}" invoked from within "catch "$className create $o $args" msg" (procedure "new" line 3) invoked from within "new [Simulator set node_factory_] $args" (procedure "_o3" line 12) (Simulator node line 12) invoked from within "$ns node" (file "rsvp_conf.tcl" line 34) ~ Does anybody know what's the matter? Thank you for reading! Ge Yu From ivanp@netgroup-serv.polito.it Fri Jan 21 08:39:30 2000 From: ivanp@netgroup-serv.polito.it (Ivan) Date: Fri, 21 Jan 2000 09:39:30 +0100 Subject: Question about CBQ Message-ID: <38881B42.78AF71CE@netgroup-serv.polito.it> Dear all, I'd tried to simulate a CBQ scheduler, this is the script: set cbqalgorithm formal set stopTime 100 set ns [new Simulator] set nf [open all.tr w] $ns trace-all $nf proc make_topology { } { global s1 r1 cbqlink ns cbqalgorithm set s1 [$ns node] set r1 [$ns node] $ns simplex-link $s1 $r1 2Mb 5ms CBQ/WRR $ns simplex-link $r1 $s1 2Mb 5ms DropTail set cbqlink [$ns link $s1 $r1] [$cbqlink classifier] resize 2000 [$cbqlink queue] algorithm $cbqalgorithm } proc make_queue { dim classe_cbq } { set q [new Queue/DropTail] $q set limit_ $dim $classe_cbq install-queue $q } proc make_class_gerarchy { } { global root_class right_class left_class cbqlink set codalim 10 set root_class [new CBQClass] set left_class [new CBQClass] set right_class [new CBQClass] $right_class set maxpkt_ 1024 $left_class set maxpkt_ 1024 $root_class set maxpkt_ 1024 $root_class setparams none 0 1.0 0 1 2 0 $left_class setparams $root_class 1 0.04 0.020 0 1 0.023 $right_class setparams $root_class 0 0.96 0.000036 1 1 0.000041 make_queue $codalim $right_class make_queue $codalim $left_class $cbqlink insert $root_class $cbqlink insert $right_class $cbqlink insert $left_class $cbqlink bind $right_class 2 $cbqlink bind $left_class 1 } proc make_source_cbr { } { global ns s1 r1 cbr_rt set cbr_rt [$ns create-connection CBR $s1 Null $r1 1] $cbr_rt set packetSize_ 120 $cbr_rt set interval_ 0.00096 } proc make_source_cbr1 { } { global ns s1 r1 cbr_rt1 set cbr_rt1 [$ns create-connection CBR $s1 Null $r1 2] $cbr_rt1 set packetSize_ 120 $cbr_rt1 set interval_ 0.00048 } proc run_sim { } { global s1 r1 cbr_rt cbr_rt1 ns stopTime $ns at 0.0 "$cbr_rt start" $ns at $stopTime "$cbr_rt stop" $ns at 0.0 "$cbr_rt1 start" $ns at $stopTime "$cbr_rt1 stop" } proc finish { } { global ns nf $ns flush-trace close $nf #exec ns post-el.tcl exec ns throuput.tcl #exec xgraph -geometry 750x500 throuput.tr teo.tr exit 0 } make_topology make_class_gerarchy make_source_cbr make_source_cbr1 run_sim $ns at [expr $stopTime+1] "finish" $ns run There are two sources: - cbr_rt1: 1Mbit/s connected to the left class (80 Kbit/s) - cbr_rt2: 2Mbit/s connected to the right class (1920 Kbit/s) Why the left class does receive a better throuput than 80Kbit/s? Thanks Ivan. From Arnaud.Legout@eurecom.fr Fri Jan 21 13:09:35 2000 From: Arnaud.Legout@eurecom.fr (Arnaud Legout) Date: Fri, 21 Jan 2000 14:09:35 +0100 Subject: bug in webtraf.cc Message-ID: <38885A8F.F7CC020A@eurecom.fr> take the file varybell.tcl choose (to shorten the simulation time): set num_server 1 set num_client 1 take the file large-scale-web-traffic.tcl choose: set numSession 1 $ns at 5000 "finish" The simulation (ns large-scale-web-traffic.tcl) will stop prematuraly with the message: done pages 250 != all pages 249 I assume the problem arose when a session does not have a page to get anymore (when the session size reaches [$sessionSize set val_ 250]), but I don't no why. Someone has an idea? Regards, Arnaud. -- ---------------------------------------------------------------------- Arnaud Legout Institut Eurecom Phone : 00.33.4.93.00.26.61 2229, route des Cretes Fax : 00.33.4.93.00.26.27 BP 193 E-mail: Arnaud.Legout@eurecom.fr 06904 Sophia Antipolis cedex Web : http://www.eurecom.fr/~legout FRANCE ---------------------------------------------------------------------- From Tarik Alj Fri Jan 21 13:46:20 2000 From: Tarik Alj (Tarik Alj) Date: Fri, 21 Jan 2000 08:46:20 -0500 (EST) Subject: Question about CBQ Message-ID: <200001211346.IAA06068@someware.INRS-Telecom.UQuebec.CA> > > There are two sources: > - cbr_rt1: 1Mbit/s connected to the left class (80 Kbit/s) > - cbr_rt2: 2Mbit/s connected to the right class (1920 Kbit/s) > > Why the left class does receive a better throuput than 80Kbit/s? I would say it's because the left class borrows from the root class; obviously it's getting (proportionnaly) way less bw than the right class. > > Thanks Ivan. > > > > Tarik From mallman@grc.nasa.gov Fri Jan 21 14:23:50 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Fri, 21 Jan 2000 09:23:50 -0500 Subject: a though to help with debugging ns Message-ID: <200001211423.JAA26826@lombok-fi.lerc.nasa.gov> > Since the variables are class variables and have local scope, why > not just be clear and consistent and call them *all* debug_ ? Nice. That would be better, I think. > I mean, all this overhyped object-oriented twaddle I keep hearing > about has to be good for _something_. Exactly. ;) allman From L.Wood@eim.surrey.ac.uk Fri Jan 21 14:39:07 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Fri, 21 Jan 2000 14:39:07 +0000 (GMT) Subject: a though to help with debugging ns In-Reply-To: <200001211423.JAA26826@lombok-fi.lerc.nasa.gov> Message-ID: On Fri, 21 Jan 2000, Mark Allman wrote: > > Since the variables are class variables and have local scope, why > > not just be clear and consistent and call them *all* debug_ ? > > Nice. That would be better, I think. Additional twist: define and bind verbose_ similarly, and #define NOW Scheduler::instance().clock() consistently across ns. That will make it easier to write self-documenting code to explain functionality and track the operation of anything you're interested in. if (verbose_) { printf("Doing foo at time %.3f\n",NOW); } and if {[$tcp0 set verbose_]} { puts "extra stuff about configuration info below" } Arguably better than comments. The difference between verbose_ and debug_ is that debug_ never gets used in release code... (How would you create a bound class variable - myname_ say,with the full name and network location of the object it belongs to, for printing out?) L. PGP From mallman@grc.nasa.gov Fri Jan 21 15:12:25 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Fri, 21 Jan 2000 10:12:25 -0500 Subject: TCP reset bug Message-ID: <200001211512.KAA04357@lombok-fi.lerc.nasa.gov> I have just run across one more problem with resetting a TCP connection (one-way TCP -- all forms, as best as I can tell). I am not using the current version of ns2, but I just quickly grepped through the daily snapshot and it looks like the bug is still present. The code works really hard to reset the RTO timer to the initial values when a TCP connection is reset. However, the rtt_seq_ variable is not reset. This can cause ns to observe a *huge* RTT (the size of which depends on how long the connection was idle for), which then bloats up the RTO. Also, I *believe* that it also prevents TCP from taking RTT samples when it really should be (because it is waiting for some segment X, where X is some large number). (I didn't spend a whole lot of time verifying that last claim). The easy fix is to set rtt_seq_ = -1 in TcpAgent::reset(). allman From guol@cs.bu.edu Fri Jan 21 15:40:45 2000 From: guol@cs.bu.edu (Guo, Liang) Date: Fri, 21 Jan 2000 10:40:45 -0500 (EST) Subject: bug in webtraf.cc In-Reply-To: <38885A8F.F7CC020A@eurecom.fr> Message-ID: On Fri, 21 Jan 2000, Arnaud Legout wrote: > take the file varybell.tcl > choose (to shorten the simulation time): > set num_server 1 > set num_client 1 > > take the file large-scale-web-traffic.tcl > choose: > set numSession 1 > $ns at 5000 "finish" > > The simulation (ns large-scale-web-traffic.tcl) will stop > prematuraly with the message: > done pages 250 != all pages 249 > > I assume the problem arose when a session does not have a page to get > anymore > (when the session size reaches [$sessionSize set val_ 250]), > but I don't no why. > > Someone has an idea? > > Regards, > > Arnaud. > > See discussion in: http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9912/0068.html and related mails therein From Hajer.Tounsi@ensi.rnu.tn Fri Jan 21 11:05:32 2000 From: Hajer.Tounsi@ensi.rnu.tn (Hajer Tounsi) Date: Fri, 21 Jan 2000 12:05:32 +0100 Subject: classifier? Message-ID: <38883D7C.286F237E@ensi.rnu.tn> Hi All, I have created a new type of Node ;mynode; which uses a new classifier ;myclassifier;. This one references a classifier Address following this scheme. Node entry -> myclassifier -> classifier address. I have defined in mynode instprc mk-default-classifier { ......... $myclassifier_ install 0 $classifier_ ......... } when I run my simulation script, I obtain this error: -o24 : no target for slot 1 could anyone help please ? hajer From pmatias@student.dei.uc.pt Fri Jan 21 15:27:46 2000 From: pmatias@student.dei.uc.pt (Pedro Nuno M Matias) Date: Fri, 21 Jan 2000 15:27:46 +0000 (WET) Subject: rtProtoLS: meaning of 'iterator' Message-ID: Hello from Portugal. I'm studing the code of your implementation rtProtoLS and, due to my inexperience in analising C++ code, I couldn't yet understand the role of the class (???) iterator. I've looked for this class (???) throught the files, and in the definition of the class LsList in ls.h I found some commented code that I suspect to be the result of the rtProtoLS development phase. Was this result of some idea of development that you didn't followed? Or I am understanding everything wrong... The word iterator seams to be vastely used in the code, for example in the definition of methods such as LsPaths::insertNextHopList, and I am totaly clueless about the meaning of it :((( Anyway I would very much appreciate that you, or anyone in the list with a better understanding of this issue than myself, could give me some answer. Obrigado e um abraco para todos. Pedro Matias ---------------------------------------- Laboratorio de Comunicacoes e Telematica Departamento de Engenharia Informatica Universidade de Coimbra - PORTUGAL http://lct.dei.uc.pt ---------------------------------------- From Tarik Alj Fri Jan 21 16:37:21 2000 From: Tarik Alj (Tarik Alj) Date: Fri, 21 Jan 2000 11:37:21 -0500 (EST) Subject: classifier? Message-ID: <200001211637.LAA06336@someware.INRS-Telecom.UQuebec.CA> > Hi All, > > > I have created a new type of Node ;mynode; which uses a new classifier > ;myclassifier;. This one references a classifier Address following this > scheme. > Node entry -> myclassifier -> classifier address. > > I have defined in mynode instprc mk-default-classifier { > ......... > $myclassifier_ install 0 $classifier_ > > ......... > } > > when I run my simulation script, I obtain this error: > > -o24 : no target for slot 1 > > could anyone help please ? > > hajer classifier_ is installed in slot 0, what's installed in slot 1 of myclassifier_? what should be? Obviously packet are sent to myclassifier_ with a dst=1 and there is nothing in the corresponding slot of myclassifier_; assuming _o24 is myclassifier_. You can see what's in a classifier by using the instproc "dump". i.e.: $classifier_ dump Tarik From Fusun Inanc Fri Jan 21 20:21:22 2000 From: Fusun Inanc (Fusun Inanc) Date: Fri, 21 Jan 2000 15:21:22 -0500 (EST) Subject: ErrorModel rate_??? Message-ID: <200001212021.PAA06936@poptop.> Thanks Lloyd, I changed it to byte and convert the bit error rate to byte error rate . but i still have the same problem. I think neither c++ nor TCL is sensitive to real numbers after 6 digits. How can I increase it? I will show my simulation results: Bit error rate = 0 Byte error rate = 0.0 Throughput for TCP connection = 7891200.0 Mbps Bit error rate = 1e-6 Byte error rate = 7.999972000294342e-06 Throughput for TCP connection = 7815360.0 Mbps Bit error rate = 1e-7 Byte error rate = 7.9999971958066851e-07 Throughput for TCP connection = 7891200.0 Mbps Simulation time 50sec. I have some dropped packets if BER is 10^-06 but if I increase it, no packet is dropped. I want to see the effect of BER=10^-08 in my simulation though. Thanks, Fusun Inanc > > I am using ErrorModel agent in my simulation. And I want to > > create a link which has a bit error rate of 1e-08. > > > > If I set the rate > > > > $loss_module1 set rate_ 1e-05 > > > > it drops packets. > > If you don't specify the error rate as anything, it's packet rate > (pkt), not bit rate. > > I think you need to translate your BER into an approximately > equivalent byte error rate, and do > > $lossmodule unit byte > > first. (the end of nsDoc 12.1 says something about "either in packets > of bits", which is extremely misleading, and imo needs fixing.) > > L. > > > > But if I set it to 1e-06 , 1e-07 and so on, It doesn't do anything. > > I run the simulation 50 sec. > > > > Is there a way to make the error rate more sensitive? > > > > Thanks in advance... > > > > Fusun Inanc > > PGP > From swtan@mmu.edu.my Sat Jan 22 06:09:15 2000 From: swtan@mmu.edu.my (Tan Su Wei) Date: Sat, 22 Jan 2000 13:09:15 +0700 Subject: Questions about ManualRtNode, hash classifier References: <200001201340.IAA04482@someware.INRS-Telecom.UQuebec.CA> Message-ID: <001801bf649f$362a9040$3b23640a@cyber.mmu.edu.my> Hi, Thanks for your reply... and i just got the ManualRtNode running... by using $n1 add-route-to-adj-node -default $r1 .. Regards Tan Su Wei ----- Original Message ----- From: Tarik Alj To: Cc: Sent: Thursday, January 20, 2000 8:40 PM Subject: Re: Questions about ManualRtNode, hash classifier > Tan, > > i think the probleme comes from: > > # manual set route > $n1 add-route-to-adj-node $r1 > $r1 add-route-to-adj-node $r2 > $r2 add-route-to-adj-node $r5 > $r5 add-route-to-adj-node $r4 > $r4 add-route-to-adj-node $n2 > > the unknown flow error comes from the classifier being unable to map the flow > from 0 to 256 with fid of 0 (1 stands for the bucket number) to a slot. Either > you are not using these intsproc correctly or you've stepped across a bug? > > > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to > owner-ns-users@irtf.CS.Berkeley.EDU using -f > > Date: Thu, 20 Jan 2000 16:21:28 +0000 > > From: Tan Su Wei > > X-Accept-Language: en > > MIME-Version: 1.0 > > To: "ns-users@mash.CS.Berkeley.EDU" > > Subject: Questions about ManualRtNode, hash classifier > > > > Dear all, > > I'd tried to simulate a manual route packet forwarding using > > ManualRtNode {in ns-node.tcl}, I got some error when trying to run a > > simple simulation script. > > I can figure out what is wrong with the script and no sure what the > > error message means. Can some one please show me how to used > > ManualRtNode or the hashclassifier for setting up a pre comfigure path. > > > > Attach is my simulation script... and below is the error message > > generated.... > > > > > ns: _o11 unknown-flow 0 256 0 1: > > (_o11 cmd line 1) > > invoked from within > > "_o11 cmd unknown-flow 0 256 0 1" > > invoked from within > > "catch "$self cmd $args" ret" > > (procedure "_o11" line 2) > > (SplitObject unknown line 2) > > invoked from within > > "_o11 unknown-flow 0 256 0 1" > > --------------------------> > > > > Regards > > Tan Su Wei > Tarik > > From syan@catarina.usc.edu Sat Jan 22 07:44:03 2000 From: syan@catarina.usc.edu (Shuqian Yan) Date: Fri, 21 Jan 2000 23:44:03 -0800 (PST) Subject: Addressing scheme in NS In-Reply-To: <38885A8F.F7CC020A@eurecom.fr> Message-ID: Hello, Can somone explain the new addressing scheme used in the latest version of ns? for example in ip.h, the following have been defined, ns_addr_t& src() { return (src_); } nsaddr_t& saddr() { return (src_.addr_); } int32_t& sport() { return src_.port_;} does ns_addr_t& src() { return (src_); } define both the node and port address for the src??? does nsaddr_t& saddr() { return (src_.addr_); } define only the node address? does int32_t& sport() { return src_.port_;} define the port address on the node?? tks. shuqian From intanago@isi.edu Sat Jan 22 07:57:20 2000 From: intanago@isi.edu (Chalermek Intanagonwiwat) Date: Fri, 21 Jan 2000 23:57:20 -0800 (PST) Subject: Addressing scheme in NS In-Reply-To: Message-ID: Yes. You got it right. ns_addr_t is a struct of nsaddr_t and int32_t. Chalermek On Fri, 21 Jan 2000, Shuqian Yan wrote: > Hello, > > Can somone explain the new addressing scheme used in the latest version > of ns? for example in ip.h, the following have been defined, > > ns_addr_t& src() { return (src_); } > nsaddr_t& saddr() { return (src_.addr_); } > int32_t& sport() { return src_.port_;} > > does ns_addr_t& src() { return (src_); } define both the node and port > address for the src??? > > does nsaddr_t& saddr() { return (src_.addr_); } define only the node > address? > > does int32_t& sport() { return src_.port_;} define the port address on the > node?? > > tks. > > shuqian > > From alhussein_abouzeid@hotmail.com Sun Jan 23 08:20:40 2000 From: alhussein_abouzeid@hotmail.com (Alhussein Abouzeid) Date: Sun, 23 Jan 2000 08:20:40 GMT Subject: exp_.max_p_inv alternating values (bug?) Message-ID: <20000123082040.36772.qmail@hotmail.com> Hi, I'm still using ns2.1b5. I'm experimenting with RED and changing linterm_, but I found that the results were not as I expected. I did a preliminary debug by printing the value of linterm_ (i.e exp_.max_p_inv) from red.cc during run-time, and noticed it alternates between its default value of 10 and its assigned value. I'm not sure how this could happen. Any suggestions? Attached is the trace result and the red.cc with the modification of printf's in different places. Thanks in advance. AA. ---------------- #trace (alternating between the default and the assigned value of 4) #the first number indicates the function from which printf is called 1 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 1 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 4.000000 7 edp_.max_p_inv is 4.000000 7 edp_.max_p_inv is 10.000000 1 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 7 edp_.max_p_inv is 4.000000 1 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 10.000000 2 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 3 edp_.max_p_inv is 10.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 3 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 5 edp_.max_p_inv is 4.000000 2 edp_.max_p_inv is 4.000000 ------------------------------------------- /*modified red.cc*/ /* -*- Mode:C++; c-basic-offset:8; tab-width:8; indent-tabs-mode:t -*- */ /* * Copyright (c) 1990-1997 Regents of the University of California. * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * 3. All advertising materials mentioning features or use of this software * must display the following acknowledgement: * This product includes software developed by the Computer Systems * Engineering Group at Lawrence Berkeley Laboratory. * 4. Neither the name of the University nor of the Laboratory may be used * to endorse or promote products derived from this software without * specific prior written permission. * * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF * SUCH DAMAGE. * * * Here is one set of parameters from one of Sally's simulations * (this is from tcpsim, the older simulator): * * ed [ q_weight=0.002 thresh=5 linterm=30 maxthresh=15 * mean_pktsize=500 dropmech=random-drop queue-size=60 * plot-file=none bytes=false doubleq=false dqthresh=50 * wait=true ] * * 1/"linterm" is the max probability of dropping a packet. * There are different options that make the code * more messy that it would otherwise be. For example, * "doubleq" and "dqthresh" are for a queue that gives priority to * small (control) packets, * "bytes" indicates whether the queue should be measured in bytes * or in packets, * "dropmech" indicates whether the drop function should be random-drop * or drop-tail when/if the queue overflows, and * the commented-out Holt-Winters method for computing the average queue * size can be ignored. * "wait" indicates whether the gateway should wait between dropping * packets. */ #ifndef lint static const char rcsid[] = "@(#) $Header: /usr/src/mash/repository/vint/ns-2/red.cc,v 1.34 1999/01/07 19:01:57 sfloyd Exp $ (LBL)"; #endif #include #include #include #include "template.h" #include "random.h" #include "flags.h" #include "delay.h" #include "red.h" static class REDClass : public TclClass { public: REDClass() : TclClass("Queue/RED") {} TclObject* create(int, const char*const*) { return (new REDQueue); } } class_red; REDQueue::REDQueue() : link_(NULL), bcount_(0), de_drop_(NULL), tchan_(0), idle_(1) { bind_bool("bytes_", &edp_.bytes); // boolean: use bytes? bind_bool("queue-in-bytes_", &qib_); // boolean: q in bytes? bind("thresh_", &edp_.th_min); // minthresh bind("maxthresh_", &edp_.th_max); // maxthresh bind("mean_pktsize_", &edp_.mean_pktsize); // avg pkt size bind("q_weight_", &edp_.q_w); // for EWMA bind_bool("wait_", &edp_.wait); bind("linterm_", &edp_.max_p_inv); bind_bool("setbit_", &edp_.setbit); // mark instead of drop bind_bool("drop-tail_", &drop_tail_); // drop last pkt bind_bool("drop-front_", &drop_front_); // drop first pkt bind_bool("drop-rand_", &drop_rand_); // drop pkt at random bind("ave_", &edv_.v_ave); // average queue sie bind("prob1_", &edv_.v_prob1); // dropping probability bind("curq_", &curq_); // current queue size q_ = new PacketQueue(); // underlying queue pq_ = q_; reset(); #ifdef notdef print_edp(); print_edv(); #endif } void REDQueue::reset() { /* * If queue is measured in bytes, scale min/max thresh * by the size of an average packet (which is specified by user). */ printf("1 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ if (qib_) { edp_.th_min *= edp_.mean_pktsize; edp_.th_max *= edp_.mean_pktsize; } /* * Compute the "packet time constant" if we know the * link bandwidth. The ptc is the max number of (avg sized) * pkts per second which can be placed on the link. * The link bw is given in bits/sec, so scale mean psize * accordingly. */ if (link_) edp_.ptc = link_->bandwidth() / (8. * edp_.mean_pktsize); edv_.v_ave = 0.0; edv_.v_slope = 0.0; edv_.count = 0; edv_.count_bytes = 0; edv_.old = 0; edv_.v_a = 1 / (edp_.th_max - edp_.th_min); edv_.v_b = - edp_.th_min / (edp_.th_max - edp_.th_min); idle_ = 1; if (&Scheduler::instance() != NULL) idletime_ = Scheduler::instance().clock(); else idletime_ = 0.0; /* sched not instantiated yet */ Queue::reset(); bcount_ = 0; } /* * Compute the average queue size. * The code contains two alternate methods for this, the plain EWMA * and the Holt-Winters method. * nqueued can be bytes or packets */ void REDQueue::run_estimator(int nqueued, int m) { printf("2 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ float f, f_sl, f_old; f = edv_.v_ave; f_sl = edv_.v_slope; #define RED_EWMA #ifdef RED_EWMA while (--m >= 1) { f_old = f; f *= 1.0 - edp_.q_w; } f_old = f; f *= 1.0 - edp_.q_w; f += edp_.q_w * nqueued; #endif #ifdef RED_HOLT_WINTERS while (--m >= 1) { f_old = f; f += f_sl; f *= 1.0 - edp_.q_w; f_sl *= 1.0 - 0.5 * edp_.q_w; f_sl += 0.5 * edp_.q_w * (f - f_old); } f_old = f; f += f_sl; f *= 1.0 - edp_.q_w; f += edp_.q_w * nqueued; f_sl *= 1.0 - 0.5 * edp_.q_w; f_sl += 0.5 * edp_.q_w * (f - f_old); #endif edv_.v_ave = f; edv_.v_slope = f_sl; } /* * Return the next packet in the queue for transmission. */ Packet* REDQueue::deque() { printf("3 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ Packet *p; p = q_->deque(); if (p != 0) { idle_ = 0; bcount_ -= ((hdr_cmn*)p->access(off_cmn_))->size(); } else { idle_ = 1; // deque() may invoked by Queue::reset at init // time (before the scheduler is instantiated). // deal with this case if (&Scheduler::instance() != NULL) idletime_ = Scheduler::instance().clock(); else idletime_ = 0.0; } return (p); } /* * should the packet be dropped/marked due to a probabilistic drop? */ int REDQueue::drop_early(Packet* pkt) { printf("4 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ hdr_cmn* ch = (hdr_cmn*)pkt->access(off_cmn_); double p = edv_.v_a * edv_.v_ave + edv_.v_b; p /= edp_.max_p_inv; edv_.v_prob1 = p; if (edv_.v_prob1 > 1.0) edv_.v_prob1 = 1.0; double count1 = edv_.count; if (edp_.bytes) count1 = (double) (edv_.count_bytes/edp_.mean_pktsize); if (edp_.wait) { if (count1 * p < 1.0) p = 0.0; else if (count1 * p < 2.0) p /= (2 - count1 * p); else p = 1.0; } else { if (count1 * p < 1.0) p /= (1.0 - count1 * p); else p = 1.0; } if (edp_.bytes && p < 1.0) { p = p * ch->size() / edp_.mean_pktsize; } if (p > 1.0) p = 1.0; edv_.v_prob = p; // drop probability is computed, pick random number and act double u = Random::uniform(); if (u <= edv_.v_prob) { // DROP or MARK edv_.count = 0; edv_.count_bytes = 0; hdr_flags* hf = (hdr_flags*)pickPacketForECN(pkt)->access(off_flags_); if (edp_.setbit && hf->ect()) { hf->ce() = 1; // mark Congestion Experienced bit return (0); // no drop } else { return (1); // drop } } return (0); // no DROP/mark } /* * Pick packet for early congestion notification (ECN). This packet is then * marked or dropped. Having a separate function do this is convenient for * supporting derived classes that use the standard RED algorithm to compute * average queue size but use a different algorithm for choosing the packet for * ECN notification. */ Packet* REDQueue::pickPacketForECN(Packet* pkt) { return pkt; /* pick the packet that just arrived */ } /* * Pick packet to drop. Having a separate function do this is convenient for * supporting derived classes that use the standard RED algorithm to compute * average queue size but use a different algorithm for choosing the victim. */ Packet* REDQueue::pickPacketToDrop() { printf("5 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ int victim; if (drop_front_) victim = min(1, q_->length()-1); else if (drop_rand_) victim = Random::integer(q_->length()); else /* default is drop_tail_ */ victim = q_->length() - 1; return(q_->lookup(victim)); } /* * Receive a new packet arriving at the queue. * The average queue size is computed. If the average size * exceeds the threshold, then the dropping probability is computed, * and the newly-arriving packet is dropped with that probability. * The packet is also dropped if the maximum queue size is exceeded. * * "Forced" drops mean a packet arrived when the underlying queue was * full or when the average q size exceeded maxthresh. * "Unforced" means a RED random drop. * * For forced drops, either the arriving packet is dropped or one in the * queue is dropped, depending on the setting of drop_tail_. * For unforced drops, the arriving packet is always the victim. */ #define DTYPE_NONE 0 /* ok, no drop */ #define DTYPE_FORCED 1 /* a "forced" drop */ #define DTYPE_UNFORCED 2 /* an "unforced" (random) drop */ void REDQueue::enque(Packet* pkt) { printf("5 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ /* * if we were idle, we pretend that m packets arrived during * the idle period. m is set to be the ptc times the amount * of time we've been idle for */ int m = 0; if (idle_) { double now = Scheduler::instance().clock(); /* To account for the period when the queue was empty. */ idle_ = 0; m = int(edp_.ptc * (now - idletime_)); } /* * Run the estimator with either 1 new packet arrival, or with * the scaled version above [scaled by m due to idle time] * (bcount_ maintains the byte count in the underlying queue). * If the underlying queue is able to delete packets without * us knowing, then bcount_ will not be maintained properly! */ run_estimator(qib_ ? bcount_ : q_->length(), m + 1); /* * count and count_bytes keeps a tally of arriving traffic * that has not been dropped (i.e. how long, in terms of traffic, * it has been since the last early drop) */ hdr_cmn* ch = (hdr_cmn*)pkt->access(off_cmn_); ++edv_.count; edv_.count_bytes += ch->size(); /* * DROP LOGIC: * q = current q size, ~q = averaged q size * 1> if ~q > maxthresh, this is a FORCED drop * 2> if minthresh < ~q < maxthresh, this may be an UNFORCED drop * 3> if (q+1) > hard q limit, this is a FORCED drop */ register double qavg = edv_.v_ave; int droptype = DTYPE_NONE; int qlen = qib_ ? bcount_ : q_->length(); int qlim = qib_ ? (qlim_ * edp_.mean_pktsize) : qlim_; curq_ = qlen; // helps to trace queue during arrival, if enabled if (qavg >= edp_.th_min && qlen > 1) { if (qavg >= edp_.th_max) { droptype = DTYPE_FORCED; } else if (edv_.old == 0) { /* * The average queue size has just crossed the * threshold from below to above "minthresh", or * from above "minthresh" with an empty queue to * above "minthresh" with a nonempty queue. */ edv_.count = 1; edv_.count_bytes = ch->size(); edv_.old = 1; } else if (drop_early(pkt)) { droptype = DTYPE_UNFORCED; } } else { /* No packets are being dropped. */ edv_.v_prob = 0.0; edv_.old = 0; } if (qlen >= qlim) { // see if we've exceeded the queue size droptype = DTYPE_FORCED; } if (droptype == DTYPE_UNFORCED) { /* pick packet for ECN, which is dropping in this case */ Packet *pkt_to_drop = pickPacketForECN(pkt); /* * If the packet picked is different that the one that just arrived, * add it to the queue and remove the chosen packet. */ if (pkt_to_drop != pkt) { q_->enque(pkt); bcount_ += ch->size(); q_->remove(pkt_to_drop); bcount_ -= ((hdr_cmn*)pkt_to_drop->access(off_cmn_))->size(); pkt = pkt_to_drop; /* XXX okay because pkt is not needed anymore */ } // deliver to special "edrop" target, if defined if (de_drop_ != NULL) de_drop_->recv(pkt); else drop(pkt); } else { /* forced drop, or not a drop: first enqueue pkt */ q_->enque(pkt); bcount_ += ch->size(); /* drop a packet if we were told to */ if (droptype == DTYPE_FORCED) { /* drop random victim or last one */ pkt = pickPacketToDrop(); q_->remove(pkt); bcount_ -= ((hdr_cmn*)pkt->access(off_cmn_))->size(); drop(pkt); } } return; } int REDQueue::command(int argc, const char*const* argv) { printf("7 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ Tcl& tcl = Tcl::instance(); if (argc == 2) { if (strcmp(argv[1], "reset") == 0) { reset(); return (TCL_OK); } if (strcmp(argv[1], "early-drop-target") == 0) { if (de_drop_ != NULL) tcl.resultf("%s", de_drop_->name()); return (TCL_OK); } } else if (argc == 3) { // attach a file for variable tracing if (strcmp(argv[1], "attach") == 0) { int mode; const char* id = argv[2]; tchan_ = Tcl_GetChannel(tcl.interp(), (char*)id, &mode); if (tchan_ == 0) { tcl.resultf("RED: trace: can't attach %s for writing", id); return (TCL_ERROR); } return (TCL_OK); } // tell RED about link stats if (strcmp(argv[1], "link") == 0) { LinkDelay* del = (LinkDelay*)TclObject::lookup(argv[2]); if (del == 0) { tcl.resultf("RED: no LinkDelay object %s", argv[2]); return(TCL_ERROR); } // set ptc now link_ = del; edp_.ptc = link_->bandwidth() / (8. * edp_.mean_pktsize); return (TCL_OK); } if (strcmp(argv[1], "early-drop-target") == 0) { NsObject* p = (NsObject*)TclObject::lookup(argv[2]); if (p == 0) { tcl.resultf("no object %s", argv[2]); return (TCL_ERROR); } de_drop_ = p; return (TCL_OK); } if (!strcmp(argv[1], "packetqueue-attach")) { delete q_; if (!(q_ = (PacketQueue*) TclObject::lookup(argv[2]))) return (TCL_ERROR); else { pq_ = q_; return (TCL_OK); } } } return (Queue::command(argc, argv)); } /* * Routine called by TracedVar facility when variables change values. * Currently used to trace values of avg queue size, drop probability, * and the instantaneous queue size seen by arriving packets. * Note that the tracing of each var must be enabled in tcl to work. */ void REDQueue::trace(TracedVar* v) { char wrk[500], *p; if (((p = strstr(v->name(), "ave")) == NULL) && ((p = strstr(v->name(), "prob")) == NULL) && ((p = strstr(v->name(), "curq")) == NULL)) { fprintf(stderr, "RED:unknown trace var %s\n", v->name()); return; } if (tchan_) { int n; double t = Scheduler::instance().clock(); // XXX: be compatible with nsv1 RED trace entries if (*p == 'c') { sprintf(wrk, "Q %g %d", t, int(*((TracedInt*) v))); } else { sprintf(wrk, "%c %g %g", *p, t, double(*((TracedDouble*) v))); } n = strlen(wrk); wrk[n] = '\n'; wrk[n+1] = 0; (void)Tcl_Write(tchan_, wrk, n+1); } return; } /* for debugging help */ void REDQueue::print_edp() { printf("8 edp_.max_p_inv is %f\n", edp_.max_p_inv); /* Hussein*/ printf("mean_pktsz: %d\n", edp_.mean_pktsize); printf("bytes: %d, wait: %d, setbit: %d\n", edp_.bytes, edp_.wait, edp_.setbit); printf("minth: %f, maxth: %f\n", edp_.th_min, edp_.th_max); printf("max_p_inv: %f, qw: %f, ptc: %f\n", edp_.max_p_inv, edp_.q_w, edp_.ptc); printf("qlim: %d, idletime: %f\n", qlim_, idletime_); printf("=========\n"); } void REDQueue::print_edv() { printf("v_a: %f, v_b: %f\n", edv_.v_a, edv_.v_b); } --------------------------------------------- Regards. AA. ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From leventi@fgan.de Mon Jan 24 09:04:21 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Mon, 24 Jan 2000 10:04:21 +0100 (MET) Subject: wireless example Message-ID: <200001240904.KAA00608@melle.ffm.fgan.de> Dear ns-users Following the ns Notes and Documentation (Page 95), I've tried to run $ns /tcl/ex/wireless.tcl and the error message I get says: You are using the tracefile format older than 1.0a5 which will not allow you to run namgraph nam: error reading last line of `out.tr' Does anybody know what I can do? Thanks in advance A.Leventi-Peetz From L.Wood@eim.surrey.ac.uk Mon Jan 24 23:06:43 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 24 Jan 2000 23:06:43 +0000 (GMT) Subject: wireless example In-Reply-To: <200001240904.KAA00608@melle.ffm.fgan.de> Message-ID: On Mon, 24 Jan 2000, Anastasia Leventi-Peetz wrote: > Following the ns Notes and Documentation (Page 95), I've tried > to run $ns /tcl/ex/wireless.tcl ~ns/ns ~ns/tcl/ex/wireless.tcl > and the error message I get says: > > You are using the tracefile format older than 1.0a5 > which will not allow you to run namgraph > nam: error reading last line of `out.tr' nam should be run on out.nam, not out.tr. If the message is correct and you're actually loading in a *nam* tracefile, then you probably need to get and build a version of nam of similar vintage to your version of ns. L. > Does anybody know what I can do? > Thanks in advance > A.Leventi-Peetz PGP From chj@lin.foa.se Mon Jan 24 20:46:59 2000 From: chj@lin.foa.se (Christian =?iso-8859-1?Q?J=F6nsson?= FOA 72) Date: Mon, 24 Jan 2000 21:46:59 +0100 Subject: IP_BROADCAST warning message !! In-Reply-To: c.christian.joensson's message of Sun, 16 Jan 2000 20:09:10 +0100. <001201bf6055$2c06e800$59a7b5d4@chj> Message-ID: <200001242046.VAA07677@arnljot.lin.foa.se> hmm, my memory served me ill, as usually, the thing I thought about was related to arp.{cc,h}, not applicable to IP_BROADCAST. I have two thoughts right now, either (i) ns should us the system's IP_BROADCAST, if it is defined, or (ii), dependent on whether ns implementations rely on the specific "numeric" value 0xffffffff or not, ns sources needs to undefine IP_BROADCAST before setting it... Here's what one might do in ip.h then. (i) #ifndef IP_BROADCAST #define IP_BROADCAST ((u_int32_t) 0xffffffff) #endif /* !IP_BROADCAST */ and (ii) #ifdef IP_BROADCAST #undef IP_BROADCAST #endif /* IP_BROADCAST */ #define IP_BROADCAST ((u_int32_t) 0xffffffff) If I may suggest something, I'd suggest approach (ii). Anyway, I haven't tried it yet... Cheers, /ChJ From floyd@aciri.org Mon Jan 24 18:49:39 2000 From: floyd@aciri.org (Sally Floyd) Date: Mon, 24 Jan 2000 10:49:39 -0800 Subject: TCP reset bug Message-ID: <200001241849.KAA86380@elk.aciri.org> >I have just run across one more problem with resetting a TCP >connection (one-way TCP -- all forms, as best as I can tell). ... >The easy fix is to set rtt_seq_ = -1 in TcpAgent::reset(). Done. Thanks for the report. If you ever had a small test to contribute to one of the validation test suites to verify the basic reset functions, I would be happy to add it in. I haven't done an exhaustive search to see if there was anything else that got missed... - Sally From yiyung@mmlab.snu.ac.kr Mon Jan 24 04:57:44 2000 From: yiyung@mmlab.snu.ac.kr (Yung Yi) Date: Mon, 24 Jan 2000 13:57:44 +0900 Subject: Mac Protocol Message-ID: <003d01bf6627$8e8b1320$e3d50198@apan.snu.ac.kr> Hi. What is the current status of ns about mac and lan protocol? I think the facts and explanations in ns manual are different from real ns release. If someone has the information about this, Please give me that information. Thanks. ------------------------------------------------------ Yung Yi Multimedia & Computer Communication Lab. Dept. of Computer Engineering, Seoul National Univ. Tel : +82-2-876-7170 Fax : +82-2-876-7171 Email : yiyung@mmlab.snu.ac.kr URL : http://mmlab.snu.ac.kr/~yiyung ------------------------------------------------------ From idwfxsx@alpha.oac.ucla.edu Mon Jan 24 19:46:24 2000 From: idwfxsx@alpha.oac.ucla.edu (idwfxsx@alpha.oac.ucla.edu) Date: Mon, 24 Jan 2000 11:46:24 -0800 (PST) Subject: negative time in trace file Message-ID: <200001241946.LAA51354@alpha.oac.ucla.edu> WHen I run mac-test.tcl using TCP protocol, I observe that the clock time is nagative in out.tr. Does anyone know what is going on here? From mallman@grc.nasa.gov Mon Jan 24 20:16:31 2000 From: mallman@grc.nasa.gov (Mark Allman) Date: Mon, 24 Jan 2000 15:16:31 -0500 Subject: TCP reset bug Message-ID: <200001242016.PAA13847@lombok-fi.lerc.nasa.gov> > Done. Thanks for the report. Sure. Here is another. It isn't really a bug, but I wanted your opinion on including it... Every once in a while when I reset a connection and start generating new data fairly soon after I see ACKs from before the reset, which really confuses things. These ACKs are perfectly legitimate (usually dup ACKs for the last packet before the reset). So, I added a conditional to the beginning of the recv() function like this: if (tcph->seqno() < 0 || tcph->seqno() > maxseq_) return; That solves my problem. I have been watching for side effects, but have not come across any yet. I was not sure if you would want to include this in ns or not, so I thought I'd run it by you. > If you ever had a small test to contribute to one of the > validation test suites to verify the basic reset functions, I > would be happy to add it in. I haven't done an exhaustive search > to see if there was anything else that got missed... OK, I'll keep that in mind... I wouldn't think it would be too hard to cook up such a script. allman From jamester@dacl3.snu.ac.kr Tue Jan 25 04:03:31 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Tue, 25 Jan 2000 13:03:31 +0900 Subject: Mcast/Control error In-Reply-To: References: Message-ID: <00012513044500.01880@jamester> I keep getting following error. Does anyone know what this mean? ns: _o427 recv -2 0: no value given for parameter "m" to "_o427" (Agent/Mcast/Control recv line 1) invoked from within "_o427 recv -2 0" Thanks -James Kwon From jamester@dacl3.snu.ac.kr Tue Jan 25 04:06:44 2000 From: jamester@dacl3.snu.ac.kr (jamester) Date: Tue, 25 Jan 2000 13:06:44 +0900 Subject: Mcast/Control error In-Reply-To: <00012513044500.01880@jamester> References: <00012513044500.01880@jamester> Message-ID: <00012513075700.01922@jamester> Dops, I forgot another error ns: _o427 recv prune 0 157: can't read "messages(157)": no such element in array (Object set line 1) invoked from within "Agent/Mcast/Control set messages($m)" (procedure "_o427" line 2) (Agent/Mcast/Control recv line 2) invoked from within "_o427 recv prune 0 157" It seems like there is something wrong with pruing process..... thanks -James Kwon From Arnaud.Legout@eurecom.fr Mon Jan 24 17:04:00 2000 From: Arnaud.Legout@eurecom.fr (Arnaud Legout) Date: Mon, 24 Jan 2000 18:04:00 +0100 Subject: bug in webtraf.cc References: Message-ID: <388C8600.5CF6839E@eurecom.fr> "Guo, Liang" wrote: > See discussion in: > http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9912/0068.html > > and related mails therein I got the latest version of webtraf.cc but I still have the error message: done pages 250 != all pages 249 Do I miss something? Arnaud. -- ---------------------------------------------------------------------- Arnaud Legout Institut Eurecom Phone : 00.33.4.93.00.26.61 2229, route des Cretes Fax : 00.33.4.93.00.26.27 BP 193 E-mail: Arnaud.Legout@eurecom.fr 06904 Sophia Antipolis cedex Web : http://www.eurecom.fr/~legout FRANCE ---------------------------------------------------------------------- From Arnaud.Legout@eurecom.fr Mon Jan 24 15:05:43 2000 From: Arnaud.Legout@eurecom.fr (Arnaud Legout) Date: Mon, 24 Jan 2000 16:05:43 +0100 Subject: [Fwd: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released] Message-ID: <388C6A47.A64421A3@eurecom.fr> This is a multi-part message in MIME format. --------------1D6A31E62DB38A7066865979 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit -- ---------------------------------------------------------------------- Arnaud Legout Institut Eurecom Phone : 00.33.4.93.00.26.61 2229, route des Cretes Fax : 00.33.4.93.00.26.27 BP 193 E-mail: Arnaud.Legout@eurecom.fr 06904 Sophia Antipolis cedex Web : http://www.eurecom.fr/~legout FRANCE ---------------------------------------------------------------------- --------------1D6A31E62DB38A7066865979 Content-Type: message/rfc822 Content-Transfer-Encoding: 7bit Content-Disposition: inline X-Mozilla-Status2: 00000000 Message-ID: <388C2374.1DF3597D@eurecom.fr> Date: Mon, 24 Jan 2000 11:03:32 +0100 From: Arnaud Legout Organization: Institut Eurecom X-Mailer: Mozilla 4.5 [en] (X11; I; SunOS 5.5.1 sun4u) X-Accept-Language: fr MIME-Version: 1.0 To: Nader Salehi Subject: Re: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released References: <200001190230.SAA22658@joy.isi.edu> Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, the download page does not seem up to date (http://www-mash.cs.berkeley.edu/ns/ns-build.html). Arnaud. -- ---------------------------------------------------------------------- Arnaud Legout Institut Eurecom Phone : 00.33.4.93.00.26.61 2229, route des Cretes Fax : 00.33.4.93.00.26.27 BP 193 E-mail: Arnaud.Legout@eurecom.fr 06904 Sophia Antipolis cedex Web : http://www.eurecom.fr/~legout FRANCE ---------------------------------------------------------------------- --------------1D6A31E62DB38A7066865979-- From n2688420@dec4000.cc.ncku.edu.tw Mon Jan 24 14:33:34 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Mon, 24 Jan 2000 22:33:34 +0800 Subject: Queue?? Message-ID: <388C62BE.91ED1C37@ccmail.ncku.edu.tw> Hi! sorry to ask a fresh man's question. i got a test file about wireless simualtion while i execute it comes the following messages : invalid command Queue/DropTail/PriQueue set Prefer_Routing_Protocols 1 but in this file it has defined the variable set opt(mac) Mac/802_11 set opt(ifq) Queue/DropTail/PriQueue set opt(ll) LL set opt(ant) Antenna/OmniAntenna . . Queue/DropTail/PriQueue set Prefer_Routing_Protocols 1 hope someone may help me build this test file or give me a hint. thanks regards From huiwenli@nortelnetworks.com Tue Jan 25 00:22:54 2000 From: huiwenli@nortelnetworks.com (Huiwen Li) Date: Mon, 24 Jan 2000 16:22:54 -0800 Subject: How can I monitor ifq in lan? Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF66CA.52E567FA Content-Type: text/plain Hi, How can I monitor ifq in a lan? I tried to use $ns attach-fmon [$ns link $n0 $Lan] $fmon where n0 is a node on the lan, and Lan is the LanNode for this lan. but it gave me the error: _o75: unable to dispatch method attach-monitors while executing "$lnk attach-monitors $isnoop $osnoop $dsnoop $fm" (procedure "_o3" line 5) (Simulator attach-fmon line 5) invoked from within "$ns attach-fmon [$ns link $n0 $lan] $ifq_fmon($i)" (procedure "build_topologyB" line 37) invoked from within "build_topologyB" (file "test91.tcl" line 795) Can anyone please give me some suggestions? Thanks a lot. Huiwen ------_=_NextPart_001_01BF66CA.52E567FA Content-Type: text/html How can I monitor ifq in lan?

Hi,

How can I monitor ifq in a lan?

I tried to use
$ns attach-fmon [$ns link $n0 $Lan] $fmon
where n0 is a node on the lan, and Lan is the LanNode for this lan.

but it gave me the error:

_o75: unable to dispatch method attach-monitors
    while executing
"$lnk attach-monitors $isnoop $osnoop $dsnoop $fm"
    (procedure "_o3" line 5)
    (Simulator attach-fmon line 5)
    invoked from within
"$ns attach-fmon [$ns link $n0 $lan] $ifq_fmon($i)"
    (procedure "build_topologyB" line 37)
    invoked from within
"build_topologyB"
    (file "test91.tcl" line 795)

Can anyone please give me some suggestions? Thanks a lot.

Huiwen

------_=_NextPart_001_01BF66CA.52E567FA-- From z_l@990.net Mon Jan 24 13:30:46 2000 From: z_l@990.net (z_l@990.net) Date: 24 Jan 2000 13:30:46 -0000 Subject: packetsize_ Message-ID: <20000124133046.18269.fmail@990.net> Dear Sir: At application level, there is an parameter:packetsize_. At RTP level, there also is such a parameter. I find the size in the tracing files. I do not know what this parameter means? Are they equal to each other? Are they embrace the overhead of protocol? Thanks Zhang lei __________________________________________________ 欢迎使用金陵热线免费电子邮件系统http://www.990.net From salehi@isi.edu Tue Jan 25 01:37:44 2000 From: salehi@isi.edu (Nader Salehi) Date: Mon, 24 Jan 2000 17:37:44 -0800 (PST) Subject: [Fwd: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released] In-Reply-To: <388C6A47.A64421A3@eurecom.fr> References: <388C6A47.A64421A3@eurecom.fr> Message-ID: <200001250137.RAA09955@boreas.isi.edu> The page should be updated tonight, that is, Pacific Time. Nader >>> "Arnaud" == Arnaud Legout writes: Arnaud> Hi, the download page does not seem up to date Arnaud> (http://www-mash.cs.berkeley.edu/ns/ns-build.html). Arnaud> Arnaud. From salehi@isi.edu Tue Jan 25 01:39:11 2000 From: salehi@isi.edu (Nader Salehi) Date: Mon, 24 Jan 2000 17:39:11 -0800 (PST) Subject: Mcast/Control error In-Reply-To: <00012513044500.01880@jamester> References: <00012513044500.01880@jamester> Message-ID: <200001250139.RAA10090@boreas.isi.edu> James, The information is inconclusive. Send me your code. Nader >>> "jamester" == jamester writes: jamester> I keep getting following error. Does anyone know what this jamester> mean? jamester> ns: _o427 recv -2 0: no value given for parameter "m" to jamester> "_o427" (Agent/Mcast/Control recv line 1) invoked from jamester> within "_o427 recv -2 0" jamester> Thanks jamester> -James Kwon From salehi@isi.edu Tue Jan 25 01:45:50 2000 From: salehi@isi.edu (Nader Salehi) Date: Mon, 24 Jan 2000 17:45:50 -0800 (PST) Subject: IP_BROADCAST warning message !! In-Reply-To: <200001242046.VAA07677@arnljot.lin.foa.se> References: <001201bf6055$2c06e800$59a7b5d4@chj> <200001242046.VAA07677@arnljot.lin.foa.se> Message-ID: <200001250145.RAA10724@boreas.isi.edu> Christian, I used the second approach to suppress the warning last weekend and will modify the source tree tomorrow. Thanks, Nader >>> "FOA" == FOA 72 writes: FOA> hmm, my memory served me ill, as usually, the thing I thought FOA> about was related to arp.{cc,h}, not applicable to IP_BROADCAST. FOA> I have two thoughts right now, either (i) ns should us the FOA> system's IP_BROADCAST, if it is defined, or (ii), dependent on FOA> whether ns implementations rely on the specific "numeric" value FOA> 0xffffffff or not, ns sources needs to undefine IP_BROADCAST FOA> before setting it... FOA> Here's what one might do in ip.h then. FOA> (i) FOA> #ifndef IP_BROADCAST #define IP_BROADCAST ((u_int32_t) FOA> 0xffffffff) #endif /* !IP_BROADCAST */ FOA> and FOA> (ii) FOA> #ifdef IP_BROADCAST #undef IP_BROADCAST #endif /* IP_BROADCAST */ FOA> #define IP_BROADCAST ((u_int32_t) 0xffffffff) FOA> If I may suggest something, I'd suggest approach (ii). FOA> Anyway, I haven't tried it yet... FOA> Cheers, FOA> /ChJ From yxh146@psu.edu Tue Jan 25 04:06:38 2000 From: yxh146@psu.edu (yingying Hou) Date: Mon, 24 Jan 2000 22:06:38 -0600 Subject: recompile problems References: <001201bf6055$2c06e800$59a7b5d4@chj><200001242046.VAA07677@arnljot.lin.foa.se> <200001250145.RAA10724@boreas.isi.edu> Message-ID: <005201bf66e9$957f8a40$74c67680@yyhou> Hi I have some problems when I recompile the ns(add new feature in flowMon,also I change the ns-default.tcl.) When I recompile the following error happens: /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0 bin/tcl-expand.tcl tcl/lib/ns-lib.tcl | ../tclcl-1.0b8/tcl2c++ et_ns_lib > gen/ns_tcl.cc sh: /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0: cannot execute sh: ../tclcl-1.0b8/tcl2c++: cannot execute *** Error code 1 make: Fatal error: Command failed for target `gen/ns_tcl.cc' child process exited abnormally After that I found that the gen/ns_tcl.cc change to 0 bytes. I have read the problem mention in following link http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9903/0102.html and i check that the makefiles line 41.It is TCLSH = /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0 tclsh8.0(size 607080) is at that directory correctly. But why these happen?Why the tclsh8.0 cannot execute? Can anybody help me? Thanks in advance. best regards Yingying Hou From swtan@mmu.edu.my Tue Jan 25 13:52:38 2000 From: swtan@mmu.edu.my (Tan Su Wei) Date: Tue, 25 Jan 2000 13:52:38 +0000 Subject: Question about ManualRtNode (again) Message-ID: <388DAAA5.E7840A52@mmu.edu.my> Hi all, I'm currently working on ns to write my own classifier... (which will much similar to hash classifier) ... So I'm looking in detail on the ManualRtNode tcl code for how to use hash classifier... I wrote a simple simulation script to test on the ManualRtNode capability of defining a pre-configure route... the script seem working only when i using the : add-route-to-adj-node [-default] node with the -default flag on... >From my understanding of the ManualRtNode Class code, the add-route-to-adj-node will call the add-route procedure : if (default flag is set ) set the class variable default_ to the next available slot (calling installNext proc) else calling hash/dest classifier procedure set-hash to install the slot in hash table... If I use the add-route-to-adj-node without the -default flag, there is an error or unknown-flow which I believe is invoked the newflow() function in classifier-hash.cc . So I'm confuse since I had set the hash entry (calling add-route-to-adj-node without -default flag), at the classify() method (call when a packet reach at a node, right?) in classifier-hash.cc, why I'll get the unknown flow error? /* from classifier-hash.cc*/ int classify(Packet * const p){ hnode *hn = lookup(p); if(hn!= NULL) // found the slot , but why always fail? return (hn->slot); else if(default_ >=0) // using default is ok return (default_); return (newflow(p)); // cause the error message!!! } And I refer to the many_tcp.tcl script (in /ns-2.1b5/ex/), seem there is a used of add-route-to-adj-node without the -default, and it seem ok when i try to run it... (and how to set the option to get nam running... ?) Besides, I'm also confuse at the codes in classifier.cc for which handle packet arrival... in method void Classfier::recv(p,h), it call the find method to return an NsObject pointer of type node. In the find method, NsObject* Classifier::find(Packet* p) { NsObject* node = NULL; /* set to null */ int cl = classify(p); /* return the slot found */ if (cl < 0 || cl >= nslot_ || (node = slot_[cl]) == 0) { // confuse??? if (default_target_) return default_target_; ... } return (node); } at the if statement, if (cl< 0 || cl >= nslot_ || (node = slot_[cl]) == 0), isn't this means that it is true only if the return value of classify() method is not in the range to the slot_ table? now, what is the default_target_? and back to the Classifier::recv method again, it will call the NsObject::recv() method to free the packet... so how it the packet know which next hop it will be forwarded to? I think there must be some misunderstanding of me about the codes, please help me clear my doubt. Thank you. Tan Su Wei Malaysia From M8702007@mail.ntust.edu.tw Tue Jan 25 08:07:11 2000 From: M8702007@mail.ntust.edu.tw (Robin) Date: Tue, 25 Jan 2000 16:07:11 +0800 Subject: help! Message-ID: <388D59AF.3F099BCD@mail.ntust.edu.tw> Hi,Sir: I have questions about how I use dmalloc with ns. First, when I make, it show me the following message: -L../dmalloc-4.3.0 -ldmalloc -L/usr/lib -ltcldbg -L../tclcl-1.0b8 -ltclcl -L../otcl-1.0a4 -lotcl -L../tkbox/lib -ltk8.0 -L../tclbox/lib -ltcl8.0 -L/usr/X11R6/lib -lXext -lX11 -ldl -lm ../dmalloc-4.3.0/libdmalloc.so: file not recognized: File format not recognized collect2: ld returned 1 exit status make: *** [ns] Error 1 Second, I need add #include "dmalloc.h" into all files. Thanks Robin Lue From Martin Koyabe Mon Jan 24 18:36:53 2000 From: Martin Koyabe (Martin Koyabe) Date: Mon, 24 Jan 2000 18:36:53 +0000 (GMT) Subject: IP_BROADCAST warning message !! Message-ID: <200001241836.SAA20519@churchward.erg.abdn.ac.uk> Nader, I have also just used the second method with the new released (ns-2.1b6) and it seems to work perfectly !! Thanks ChJ !! Ofcourse ~/ns-2.1b6/ip.h needs updating !! -- Martin > Date: Mon, 24 Jan 2000 17:45:50 -0800 (PST) > From: Nader Salehi > MIME-Version: 1.0 > Content-Transfer-Encoding: 7bit > To: Christian J鰊sson FOA 72 > Cc: "Martin Koyabe" , ns-users@mash.CS.Berkeley.EDU > Subject: Re: IP_BROADCAST warning message !! > X-URL: http://www.isi.edu/~salehi > > Christian, > > I used the second approach to suppress the warning last weekend and > will modify the source tree tomorrow. > > Thanks, > Nader > > >>> "FOA" == FOA 72 writes: > > FOA> hmm, my memory served me ill, as usually, the thing I thought > FOA> about was related to arp.{cc,h}, not applicable to IP_BROADCAST. > > FOA> I have two thoughts right now, either (i) ns should us the > FOA> system's IP_BROADCAST, if it is defined, or (ii), dependent on > FOA> whether ns implementations rely on the specific "numeric" value > FOA> 0xffffffff or not, ns sources needs to undefine IP_BROADCAST > FOA> before setting it... > > FOA> Here's what one might do in ip.h then. > > FOA> (i) > > FOA> #ifndef IP_BROADCAST #define IP_BROADCAST ((u_int32_t) > FOA> 0xffffffff) #endif /* !IP_BROADCAST */ > > FOA> and > > FOA> (ii) > > FOA> #ifdef IP_BROADCAST #undef IP_BROADCAST #endif /* IP_BROADCAST */ > FOA> #define IP_BROADCAST ((u_int32_t) 0xffffffff) > > FOA> If I may suggest something, I'd suggest approach (ii). > > FOA> Anyway, I haven't tried it yet... > > FOA> Cheers, > > FOA> /ChJ > > > > From Martin Koyabe Mon Jan 24 18:40:35 2000 From: Martin Koyabe (Martin Koyabe) Date: Mon, 24 Jan 2000 18:40:35 +0000 (GMT) Subject: [Fwd: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released] Message-ID: <200001241840.SAA20752@churchward.erg.abdn.ac.uk> Nader, Remember to up date the 'ns-users Jan'2000 archives' page as well http://www-mash.cs.berkeley.edu/dist/archive/ns-users/0001 Last update 14 Jan 2000 !! -- Martin > > The page should be updated tonight, that is, Pacific Time. > > Nader > > >>> "Arnaud" == Arnaud Legout writes: > > Arnaud> Hi, the download page does not seem up to date > Arnaud> (http://www-mash.cs.berkeley.edu/ns/ns-build.html). > > Arnaud> Arnaud. > --------------------------------------------------------------------- Martin L.W.D Koyabe e-mail: koyabe@erg.abdn.ac.uk Electronics Research Group Fraser Noble Building King's College, AB24 3UE (Off) +44-01224-272813 Aberdeen, UK (Mob) +44-07881-610825 From L.Wood@eim.surrey.ac.uk Tue Jan 25 11:26:38 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Tue, 25 Jan 2000 11:26:38 +0000 (GMT) Subject: [Fwd: NS-2.1b6, Nam-1.0a8 and ns-allinone-2.1b6 were released] In-Reply-To: <200001241840.SAA20752@churchward.erg.abdn.ac.uk> Message-ID: On Mon, 24 Jan 2000, Martin Koyabe wrote: > Nader, > > Remember to up date the 'ns-users Jan'2000 archives' page as well > > http://www-mash.cs.berkeley.edu/dist/archive/ns-users/0001 > > Last update 14 Jan 2000 !! Well, half the messages for December 1999 are not visible from its index (different hypermail versions generated the index and the email pages - bad idea) and the rest of the archive is lossy, as I pointed out in a previous mail to the list - that I can't point you at because archival hasn't reached it yet. L. wonders if hypermail could be running on a cron job. PGP From Gwendal.Le-Grand@lip6.fr Tue Jan 25 13:36:06 2000 From: Gwendal.Le-Grand@lip6.fr (Gwendal LE GRAND) Date: Tue, 25 Jan 2000 14:36:06 +0100 Subject: Hierarchical adressing Message-ID: Hello I want to simulate the following architecture #Wired domain # HA--------------- FA the home and foreign agents # | lan3 | # Wless lan1 | |wireless lan2 # All the nodes # When i call make-lan, what should the address of lan3 be ? Here we have 2 domains (under HA and FA) with one cluster in each then i do set lan1_ [$ns make-lan ... -address "0.0.0" ] set lan2_ [$ns make-lan ... -address "1.0.0"] But what about lan3 (does it also have a -adress option ?) Thanks in advance GWendal -------Gwendal LE GRAND------- mailto:Gwendal.Le-Grand@lip6.fr tel: +33 (0) 1 44 27 75 12 http://www-rp.lip6.fr/~legrand fax: +33 (0) 1 44 27 74 95 Universite Pierre et Marie Curie, Laboratoire LIP6-CNRS, 8 Rue du Capitaine Scott, 75015 Paris, France From Tarik Alj Tue Jan 25 13:40:43 2000 From: Tarik Alj (Tarik Alj) Date: Tue, 25 Jan 2000 08:40:43 -0500 (EST) Subject: Mac Protocol Message-ID: <200001251340.IAA01406@someware.INRS-Telecom.UQuebec.CA> The architecture described in the notes (manual) is still valid : LANs are constructed as virtual nodes. The stuff on CSMA/CD is not. The mac classifier is not used. mac-802.3 has been updated and is now supposed to work alright. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > From: "Yung Yi" > To: > Subject: Mac Protocol > Date: Mon, 24 Jan 2000 13:57:44 +0900 > MIME-Version: 1.0 > X-Priority: 3 > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 > Content-Transfer-Encoding: 8bit > X-MIME-Autoconverted: from base64 to 8bit by mash.CS.Berkeley.EDU id PAA00736 > > Hi. > > What is the current status of ns about mac and lan protocol? > > I think the facts and explanations in ns manual are different from real ns release. > If someone has the information about this, Please give me that information. > > Thanks. > ------------------------------------------------------ > Yung Yi > Multimedia & Computer Communication Lab. > Dept. of Computer Engineering, Seoul National Univ. > > Tel : +82-2-876-7170 > Fax : +82-2-876-7171 > Email : yiyung@mmlab.snu.ac.kr > URL : http://mmlab.snu.ac.kr/~yiyung > ------------------------------------------------------ > Tarik From Tan H P" This is a multi-part message in MIME format. ------=_NextPart_000_000C_01BF6781.92B46040 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi everyone,=20 I have 2 questions about ns.=20 (a) Sometimes, I get the message that ns scheduler is going back in time = from t1 to t2. Does anyone know what is the cause of this? (b) When I have many packets in the simulation, the simulation 'hangs' = sometimes, and ns does not exit normally. Is there a memory limitation = in ns? I would appreciate any reply very very much. rgds HP Tan=20 ------=_NextPart_000_000C_01BF6781.92B46040 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi everyone,

I have 2 questions about ns.
 
(a) Sometimes, I get the message that ns scheduler is going back in = time=20 from t1 to t2. Does anyone know what is the cause of this?
 
(b) When I have many packets in the simulation, the simulation = 'hangs'=20 sometimes, and ns does not exit normally. Is there a memory limitation = in=20 ns?
 
I would appreciate any reply very very much.
 
rgds
HP Tan 
------=_NextPart_000_000C_01BF6781.92B46040-- From x Hung" Hi, I'm a fresh user of ns-2. I want to install ns-allinone-2.1b6. But there are problems. I tried ns-allinone-2.1b5 too. The same problem was encountered. Can anyone help me? Thank you very much in advance. The OS is sparc-sun-solaris2.7 Thanks, X Hung ----------- ...... cc -c -O -I./../generic -I. -DHAVE_GETCWD=1 -DHAVE_UNISTD_H=1 -DUSE_TERMIO S=1 -DHAVE_SYS_TIME_H=1 -DTIME_WITH_SYS_TIME=1 -DHAVE_TZNAME=1 -DHAVE_TIMEZONE_V AR=1 -DHAVE_ST_BLKSIZE=1 -DSTDC_HEADERS=1 -DNO_UNION_WAIT=1 -DNEED_MATHERR=1 -DRETSIGTYPE=void -DHAVE_SYS_IOCTL_H=1 -DHAVE_SYS_FILIO_H=1 -DTCL_SHLIB_EXT=\".so\" ./../generic/tclIOSock.c cc -c -O -I./../generic -I. -DHAVE_GETCWD=1 -DHAVE_UNISTD_H=1 -DUSE_TERMIO S=1 -DHAVE_SYS_TIME_H=1 -DTIME_WITH_SYS_TIME=1 -DHAVE_TZNAME=1 -DHAVE_TIMEZONE_V AR=1 -DHAVE_ST_BLKSIZE=1 -DSTDC_HEADERS=1 -DNO_UNION_WAIT=1 -DNEED_MATHERR=1 -DRETSIGTYPE=void -DHAVE_SYS_IOCTL_H=1 -DHAVE_SYS_FILIO_H=1 -DTCL_SHLIB_EXT=\".so\" ./../generic/tclIOUtil.c "./../generic/tclIOUtil.c", line 66: undefined symbol: access "./../generic/tclIOUtil.c", line 66: warning: initialization type mismatch "./../generic/tclIOUtil.c", line 66: non-constant initializer: op "U&" "./../generic/tclIOUtil.c", line 242: cannot recover from previous errors cc: acomp failed for ./../generic/tclIOUtil.c make: *** [tclIOUtil.o] Error 2 tcl8.0.4 make failed! Exiting ... From mark_wibble@hotmail.com Tue Jan 25 14:57:58 2000 From: mark_wibble@hotmail.com (Mark Butler) Date: Tue, 25 Jan 2000 06:57:58 PST Subject: Sending application data over SRM Message-ID: <20000125145758.3954.qmail@hotmail.com> I want to use NS to send application data using a multicast protocol. I would like to use SRM. However, looking in the manual, SRM has no provision for receiving application data. But it then goes on to say that the agent asumes that any application data is placed in the data portion of the packet, pointed to by packet->accessdata(). Does this mean that I can directly send application data using the above pointer ? Or do I have to create an application wrapper for the SRM agent ? Thanks, Mark ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From leventi@fgan.de Tue Jan 25 15:41:36 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Tue, 25 Jan 2000 16:41:36 +0100 (MET) Subject: HELP! Message-ID: <200001251541.QAA17171@melle.ffm.fgan.de> Dear all, I have installed the ns-2.1b6 from the pieces and also nam, all versions of January 19. When I try the examples, any examples, I get the message You are using the tracefile format older than 1.0a5 which will not allow you to run namgraph nam: badly formatted event at offset 38 in /myhome/work/out.tr nam: `+ 1 0 2 cbr 210 ------- 0 0.0 3.1 0 0' does anybody know what the problem is?? thanks A.Leventi-Peetz From ana_delgado77@hotmail.com Tue Jan 25 15:58:15 2000 From: ana_delgado77@hotmail.com (Ana Delgado) Date: Tue, 25 Jan 2000 15:58:15 GMT Subject: Multicast and Mobile IP Message-ID: <20000125155815.42565.qmail@hotmail.com> Hi everyone, I'm trying to define a network where I can use Mobile IP and some of the nodes can belong to a Multicast Group at the same time. The format of the addresses has to be changed to hierarchical if I want to run Mobile IP. And at the same time I have to set up multicast requirements that afect to the addresses of the nodes. Does anyone know if both things are incompatible and if they're not, how can I define the addresses? Thanks in advance, Ana ________________________________ Ana Delgado Centre for Telecommunications Research King's College London London England ana_delgado77@hotmail.com ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From Hajer.Tounsi@ensi.rnu.tn Tue Jan 25 16:28:21 2000 From: Hajer.Tounsi@ensi.rnu.tn (Hajer Tounsi) Date: Tue, 25 Jan 2000 17:28:21 +0100 Subject: classifier? References: <200001211637.LAA06336@someware.INRS-Telecom.UQuebec.CA> Message-ID: <388DCF25.3D2568F2@ensi.rnu.tn> Hi, > > Hi All, > > > > > > I have created a new type of Node ;mynode; which uses a new classifier > > ;myclassifier;. This one references a classifier Address following this > > scheme. > > Node entry -> myclassifier -> classifier address. > > > > I have defined in mynode instprc mk-default-classifier { > > ......... > > $myclassifier_ install 0 $classifier_ > > > > ......... > > } > > > > when I run my simulation script, I obtain this error: > > > > -o24 : no target for slot 1 > > > > could anyone help please ? > > > > hajer > > classifier_ is installed in slot 0, what's installed in slot 1 of >myclassifier_? what should be? nothing is installed in slot 1. _o24 is referencing classifier_ > You can see what's in a classifier by using the instproc "dump". i.e.: > $classifier_ dump I have done that, I think I should precise which are the target of the classifier_ , but I don't know where? classifier_ is a Classifier/Addr.I have created it in the mk-default-classifier of mynode. The entry of mynode is myclassifier_ any idee? should I precise something? hajer From Martin Koyabe Tue Jan 25 02:21:56 2000 From: Martin Koyabe (Martin Koyabe) Date: Tue, 25 Jan 2000 02:21:56 +0000 (GMT) Subject: HELP! Message-ID: <200001250221.CAA25617@churchward.erg.abdn.ac.uk> Anastasia, I suspect your problem is NOT the installation but the file you are trying to load/run in NAM. NAM will NOT understand the '.tr' file unless you use some script to convert ns-tracefile to nam file. However, to see how nam works in your tcl script, try add this in the body of your tcl file !! ---------- . . # create the nam output file set nf [open test.nam w] $ns namtrace-all $nf . . In procedure finish remember to closer the file with . . proc finish { } { global ns nf $ns flush-trace close $nf . . exit 0 } ------ Once the simulation is done, try loading file in nam % nam test.nam Good luck -- Martin p/s Check Greis's Tutorial examples for more details !! or try loading the files in ~/nam-1.0a8/ex/*.nam in NAM to see how nam works !! > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to > which will not allow you to run namgraph > nam: badly formatted event at offset 38 in /myhome/work/out.tr > nam: `+ 1 0 2 cbr 210 ------- 0 0.0 3.1 0 0' > > does anybody know what the problem is?? > > thanks > A.Leventi-Peetz From n2688420@dec4000.cc.ncku.edu.tw Tue Jan 25 17:15:08 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Wed, 26 Jan 2000 01:15:08 +0800 Subject: snapshot's problem?? Message-ID: <388DDA1B.A6C6B783@ccmail.ncku.edu.tw> Hi! for another new hand's question. If i start to update the daily snapshot, should I update all the snapshot (from 1999 - 2000), or just the latest one?? else while getting into trouble i go for updating first? and if i finish updating, can I delete those files or as cmu's wireless utilities stiil stay there for use. another problem happened on my update (1/25) of NS-2.1b5 on my Lynux. the following is the error message. how should i do to solve this problem?? Have somebody meet this problem?? object.cc: In method `int NsObject::delay_bind_dispatch(const char *, const char *, class TclObject *)': object.cc:73: no matching function for call to `TclObject::delay_bind_dispatch ( const char *&, const char *&, TclObject *&)' /usr/local/src/tclcl-1.0b8/tclcl.h:214: candidates are: TclObject::delay_bind_di spatch(const char *, const char *) make: *** [object.o] Error 1 thanks for everyone' advise regards From haoboy@isi.edu Tue Jan 25 18:22:14 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 25 Jan 2000 10:22:14 -0800 (PST) Subject: recompile problems In-Reply-To: <005201bf66e9$957f8a40$74c67680@yyhou> Message-ID: Can you go to those directories and check that tclsh8.0 and tcl2c++ can indeed be executed? - Haobo On Mon, 24 Jan 2000, yingying Hou wrote: > Hi > > I have some problems when I recompile the ns(add new feature in > flowMon,also I change the ns-default.tcl.) > When I recompile the following error happens: > > /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0 bin/tcl-expand.tcl > tcl/lib/ns-lib.tcl | ../tclcl-1.0b8/tcl2c++ et_ns_lib > gen/ns_tcl.cc > sh: /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0: cannot execute > sh: ../tclcl-1.0b8/tcl2c++: cannot execute > *** Error code 1 > make: Fatal error: Command failed for target `gen/ns_tcl.cc' > child process exited abnormally > > After that I found that the gen/ns_tcl.cc change to 0 bytes. > > I have read the problem mention in following link > http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9903/0102.html > and i check that the makefiles line 41.It is > TCLSH = /export/home/ns2/ns-allinone-2.1b5/tclbox/bin/tclsh8.0 > tclsh8.0(size 607080) is at that directory correctly. > But why these happen?Why the tclsh8.0 cannot execute? > Can anybody help me? > Thanks in advance. > > best regards > Yingying Hou > > From huiwenli@nortelnetworks.com Tue Jan 25 19:05:31 2000 From: huiwenli@nortelnetworks.com (Huiwen Li) Date: Tue, 25 Jan 2000 11:05:31 -0800 Subject: snapshot's problem?? Message-ID: This message is in MIME format. Since your mail reader does not understand this format, some or all of this message may not be legible. ------_=_NextPart_001_01BF6767.269C4572 Content-Type: text/plain I met this problem before. You need to update your tclcl-1.0b8 to daily snapshot too. Should use the latest one. Hope it helps. Huiwen > -----Original Message----- > From: CWen56 [SMTP:n2688420@dec4000.cc.ncku.edu.tw] > Sent: Tuesday, January 25, 2000 9:15 AM > To: ns-users > Subject: snapshot's problem?? > > Hi! > for another new hand's question. If i start to update the daily > snapshot, should I update all the snapshot (from 1999 - 2000), or just > the latest one?? else while getting into trouble i go for updating > first? and if i finish updating, can I delete those files or as cmu's > wireless utilities stiil stay there for use. > another problem happened on my update (1/25) of NS-2.1b5 on my > Lynux. the following is the error message. how should i do to solve this > problem?? Have somebody meet this problem?? > > object.cc: In method `int NsObject::delay_bind_dispatch(const char *, > const char > *, class TclObject *)': > object.cc:73: no matching function for call to > `TclObject::delay_bind_dispatch ( > const char *&, const char *&, TclObject *&)' > /usr/local/src/tclcl-1.0b8/tclcl.h:214: candidates are: > TclObject::delay_bind_di > spatch(const char *, const char *) > make: *** [object.o] Error 1 > > thanks for everyone' advise > regards > ------_=_NextPart_001_01BF6767.269C4572 Content-Type: text/html Content-Transfer-Encoding: quoted-printable RE: snapshot's problem??

I met this problem = before. You need to update your tclcl-1.0b8 to daily snapshot too. = Should use the latest one. Hope it helps.

Huiwen

    -----Original Message-----
    From:   CWen56 = [SMTP:n2688420@dec4000.cc.ncku.edu.tw]
    Sent:   Tuesday, January 25, 2000 9:15 AM
    To:     ns-users
    Subject:       = snapshot's problem??

    Hi!
        for another new = hand's question. If i start to update the daily
    snapshot, should I update all the = snapshot (from 1999 - 2000), or just
    the latest one?? else while getting = into trouble i go for updating
    first? and if i finish updating, = can I delete those files or as cmu's
    wireless utilities stiil stay there = for use.
        another problem = happened on my update (1/25) of NS-2.1b5 on my
    Lynux. the following is the error = message. how should i do to solve this
    problem?? Have somebody meet this = problem??

    object.cc: In method `int = NsObject::delay_bind_dispatch(const char *,
    const char
     *, class TclObject = *)':
    object.cc:73: no matching function = for call to
    `TclObject::delay_bind_dispatch = (
    const char *&, const char = *&, TclObject *&)'
    /usr/local/src/tclcl-1.0b8/tclcl.h:214: candidates = are:
    TclObject::delay_bind_di
    spatch(const char *, const char = *)
    make: *** [object.o] Error 1

    thanks for everyone' advise
    regards

------_=_NextPart_001_01BF6767.269C4572-- From puginifabio@tiscalinet.it Tue Jan 25 18:57:56 2000 From: puginifabio@tiscalinet.it (Fabio Pugini) Date: Tue, 25 Jan 2000 19:57:56 +0100 Subject: Installation help Message-ID: <001001bf6767$81de86c0$6b7a0b3e@default> This is a multi-part message in MIME format. ------=_NextPart_000_0004_01BF676E.790FB800 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi=20 While installing all-in-one ns on a Red-Hat-6.1-computer I found this = message: cc: Internal compiler error:program cc got fatal signal 11 make : ***[tcl Execute.o]error 1 tcl 8.0.4 make failed! exiting... Is this error similar to the one explained in the on-line istallation = guide about [tclPosixSt.o]? If so, what changes have to be done, and = where? Thanks. ------=_NextPart_000_0004_01BF676E.790FB800 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
While installing all-in-one ns on a = Red-Hat-6.1-computer I=20 found this message:
 
cc: Internal compiler error:program cc got fatal = signal=20 11
make : ***[tcl Execute.o]error = 1
tcl 8.0.4 make failed! = exiting...
 
Is this error similar to the one explained in the = on-line=20 istallation guide about [tclPosixSt.o]? If so, what changes have to be = done, and=20 where?
Thanks.
 
------=_NextPart_000_0004_01BF676E.790FB800-- From Fusun Inanc Tue Jan 25 20:03:27 2000 From: Fusun Inanc (Fusun Inanc) Date: Tue, 25 Jan 2000 15:03:27 -0500 (EST) Subject: to use ns in a shared directory Message-ID: <200001252003.PAA11632@poptop.> Hi everyone, I have installed ns-allinone-2.1b5 on to my Sun Solaris 2.7 unix server. In my lab, three students are using ns for their projects. I cannot install ns in three different directories because I don't have this much space in the server. And also later on the number of students who use ns might increase . Instead we all have to use the same source of ns. For this purpose I created a common user in my unix system and installed ns in to this user's home directory. and give permission to the group to use these files. But this cause the problem. Everytime whenever one of us compiles ns, the permission of some files is changed to that user as wrx and group permission is gone. And whenever the other user wants to compile it s/he has problems like permission denied... How can I solve this problem. Any help on this will be greatly appreciated. Thanks in advance. Fusun Inanc From Yunjung Yi" This is a multi-part message in MIME format. ------=_NextPart_000_0005_01BF672D.61BF27D0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Why do you use virtual base classes in TCP protocols : Reno, NewReno, = Vegas? Pls, reply to me..=20 thanks. ------=_NextPart_000_0005_01BF672D.61BF27D0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Why do you use virtual base classes in = TCP=20 protocols : Reno, NewReno, Vegas?
Pls, reply to me..
thanks.
------=_NextPart_000_0005_01BF672D.61BF27D0-- From L.Wood@eim.surrey.ac.uk Tue Jan 25 20:37:00 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Tue, 25 Jan 2000 20:37:00 +0000 (GMT) Subject: to use ns in a shared directory In-Reply-To: <200001252003.PAA11632@poptop.> Message-ID: On Tue, 25 Jan 2000, Fusun Inanc wrote: > I have installed ns-allinone-2.1b5 on to my Sun Solaris 2.7 unix server. > In my lab, three students are using ns for their projects. I cannot install > ns in three different directories because I don't have this much space in the > server. And also later on the number of students who use ns might increase . > Instead we all have to use the same source of ns. > > For this purpose I created a common user in my unix system and installed ns in > to this user's home directory. and give permission to the group to use these > files. > > But this cause the problem. Everytime whenever one of us compiles ns, the > permission of some files is changed to that user as wrx and group permission is > gone. And whenever the other user wants to compile it s/he has problems like > permission denied... > > How can I solve this problem. Any help on this will be greatly appreciated. Write a find script that runs through everywhere changing permissions, and run it before/after each compile. Something like: #!/bin/csh # I really should read 'Csh Programming Considered Harmful' again. # Quicker, darker, more seductive it is. not more powerful. # modify to suit setenv sharednsdir /user/shareduser/ns-2.1b5 # do the directories find $sharednsdir -type d -print -exec chmod 770 {} \; # do the files find $sharednsdir -type f -print -exec chmod 770 {} \; ...although you'll probably want to enlarge on and fine-tune this for various filetypes/directories on -name pattern-match to minimise the potential for editing screwups, and in any case you should really be getting your students to check their changes into a cvs tree held elsewhere, since major uncoordinated revisions by multiple people sans source control is a disaster waiting to happen. L. 'hey, my simulation just died!' 'oh, I started recompiling. so sorry.' oh yeah, chmod the binary to read-only so it must be explicitly unset. PGP From haoboy@isi.edu Tue Jan 25 20:38:27 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 25 Jan 2000 12:38:27 -0800 (PST) Subject: to use ns in a shared directory In-Reply-To: <200001252003.PAA11632@poptop.> Message-ID: I have a perl script called dup for this. You can find it at http://www.isi.edu/~haoboy/files/dup.tar.gz Let me know if you have questions. - Haobo On Tue, 25 Jan 2000, Fusun Inanc wrote: > Hi everyone, > > I have installed ns-allinone-2.1b5 on to my Sun Solaris 2.7 unix server. > In my lab, three students are using ns for their projects. I cannot install > ns in three different directories because I don't have this much space in the > server. And also later on the number of students who use ns might increase . > Instead we all have to use the same source of ns. > > For this purpose I created a common user in my unix system and installed ns in > to this user's home directory. and give permission to the group to use these > files. > > But this cause the problem. Everytime whenever one of us compiles ns, the > permission of some files is changed to that user as wrx and group permission is > gone. And whenever the other user wants to compile it s/he has problems like > permission denied... > > How can I solve this problem. Any help on this will be greatly appreciated. > > Thanks in advance. > > Fusun Inanc > > From haoboy@isi.edu Tue Jan 25 20:41:09 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 25 Jan 2000 12:41:09 -0800 (PST) Subject: to use ns in a shared directory In-Reply-To: <200001252003.PAA11632@poptop.> Message-ID: Oops, I forgot to explain. dup is used to make symbolic links to the same source directory and it allows each user to build in their own directories and have their own source modifications to be linked in. See its README for more details. - Haobo From wongcne@singnet.com.sg Tue Jan 25 21:47:50 2000 From: wongcne@singnet.com.sg (Eric Wong) Date: Wed, 26 Jan 2000 05:47:50 +0800 Subject: How to manual configure 2 separate paths? References: <388DAAA5.E7840A52@mmu.edu.my> Message-ID: <004d01bf677d$d4e4ca80$973e8489@ericwong> This is a multi-part message in MIME format. ------=_NextPart_000_004A_01BF67C0.E1C921E0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Hi , I have problem defining separate paths for each pair of source and destination nodes. Also it doesn't seems to work for in opposite direction using the same path. I only managed to get sources n1 & m1 to aggregate together and transverse via path 1 to n1. Not even n1-n2 & m1-m2 using the same path. Am I going wrong some where? I am attaching my codes. I really appreciate if you can take a look at it. After going through Chapter 20 Unicast Routing, I am not too sure how this command works? Can some one please explain? $node add-routes esp. the argument. Any comments is greatly appreciated. Thanks. Eric ------=_NextPart_000_004A_01BF67C0.E1C921E0 Content-Type: application/octet-stream; name="MyPaths.tcl" Content-Transfer-Encoding: 7bit Content-Disposition: attachment; filename="MyPaths.tcl" #simulator instance set ns [new Simulator] $ns color 1 blue $ns color 2 red #Open file for nam trace file set nf [open out.nam w] $ns namtrace-all $nf #Finish procedure proc finish {} { global ns nf $ns flush-trace #close trace file close $nf #execute nam exec nam out1.nam & exit 0 } #define topology # # n1 r2--r5 n2 # \ / \ / # r1 r4 # / \ / \ # m1 \_r3_/ m2 # # path 1 n1-r1-r2-r5-r4-n2 # path 2 m1-r1-r3-r4-m2 $ns rtproto Manual #Define nodes set n1 [$ns node] set n2 [$ns node] set m1 [$ns node] set m2 [$ns node] set r1 [$ns node] set r2 [$ns node] set r3 [$ns node] set r4 [$ns node] set r5 [$ns node] #Define links $ns duplex-link $n1 $r1 1Mb 10ms DropTail $ns duplex-link $r1 $r2 1Mb 10ms DropTail $ns duplex-link $r2 $r5 1Mb 10ms DropTail $ns duplex-link $r1 $r3 1Mb 10ms DropTail $ns duplex-link $r5 $r4 1Mb 10ms DropTail $ns duplex-link $r3 $r4 1Mb 10ms DropTail $ns duplex-link $r4 $n2 1Mb 10ms DropTail # manual set path 1 $n1 add-route-to-adj-node -default $r1 $r1 add-route-to-adj-node -default $r2 $r2 add-route-to-adj-node -default $r5 $r5 add-route-to-adj-node -default $r4 $r4 add-route-to-adj-node -default $n2 # manual set path 2 $m1 add-route-to-adj-node -default $r1 $r1 add-route-to-adj-node -default $r3 $r3 add-route-to-adj-node -default $r4 $r4 add-route-to-adj-node -default $m2 #Define agent set udp1 [new Agent/UDP] $ns attach-agent $n1 $udp1 $udp1 fid_ 1 set cbr1 [new Application/Traffic/CBR] $cbr1 set packetSize_ 100 $cbr1 set rate_ 1Mb $cbr1 attach-agent $udp1 set udp2 [new Agent/UDP] $ns attach-agent $m1 $udp2 $udp2 fid_ 2 set cbr2 [new Application/Traffic/CBR] $cbr2 set packetSize_ 100 $cbr2 set rate_ 1Mb $cbr2 attach-agent $udp2 set sink1 [new Agent/Null] $ns attach-agent $n2 $sink1 $ns connect $udp1 $sink1 set sink2 [new Agent/Null] $ns attach-agent $n2 $sink2 $ns connect $udp2 $sink2 # event $ns at 0.05 "$cbr1 start" $ns at 0.10 "$cbr2 start" $ns at 5.0 "$cbr1 stop" $ns at 4.0 "$cbr2 stop" $ns at 5.05 "finish" $ns run ------=_NextPart_000_004A_01BF67C0.E1C921E0-- From yang@sce.carleton.ca Tue Jan 25 22:10:47 2000 From: yang@sce.carleton.ca (Jiangbin Yang) Date: Tue, 25 Jan 2000 17:10:47 -0500 Subject: ns files not found on the server Message-ID: <388E1F67.531D403A@sce.carleton.ca> Hello, The files (e.g., the download file for ns-2.1b6) in http://www-mash.CS.Berkeley.EDU/ns/CHANGES.html could not be downloaded. Below is the error msg returned: > Not Found > The requested URL /dist/ns-src-2.1b6.tar.gz was not found on this server. > Apache/1.3.9 Server at www-mash.CS.Berkeley.EDU Port 80 Could you please check this? Thanks, Jiangbin Yang ______________________ Broadband Networks Lab Dept. of Systems & Computer Engineering Carleton University Ottawa, Ontario K1S 5B6, Canada Tel: (613) 520-2600 x2913 Fax: (613) 520-5727 http://www.sce.carleton.ca/~yang ________________________________ From haoboy@isi.edu Tue Jan 25 22:25:38 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 25 Jan 2000 14:25:38 -0800 (PST) Subject: ns files not found on the server In-Reply-To: <388E1F67.531D403A@sce.carleton.ca> Message-ID: Yes the link is indeed broken. But CHANGES.html is included in distributions and snapshots, so for now you can take a look there. We'll fix it soon. - Haobo On Tue, 25 Jan 2000, Jiangbin Yang wrote: > Hello, > > The files (e.g., the download file for ns-2.1b6) in > http://www-mash.CS.Berkeley.EDU/ns/CHANGES.html > could not be downloaded. Below is the error msg returned: > > > Not Found > > > The requested URL /dist/ns-src-2.1b6.tar.gz was not found on this server. > > > > Apache/1.3.9 Server at www-mash.CS.Berkeley.EDU Port 80 > > Could you please check this? > > Thanks, > > Jiangbin Yang > ______________________ > Broadband Networks Lab > Dept. of Systems & Computer Engineering > Carleton University > Ottawa, Ontario K1S 5B6, Canada > Tel: (613) 520-2600 x2913 > Fax: (613) 520-5727 > http://www.sce.carleton.ca/~yang > ________________________________ > From salehi@isi.edu Tue Jan 25 22:25:41 2000 From: salehi@isi.edu (Nader Salehi) Date: Tue, 25 Jan 2000 14:25:41 -0800 (PST) Subject: ns files not found on the server In-Reply-To: <388E1F67.531D403A@sce.carleton.ca> References: <388E1F67.531D403A@sce.carleton.ca> Message-ID: <200001252225.OAA02661@joy.isi.edu> Jiangbin, The URL in CHANGES.html is wrong. The actual URL is http://www-mash.cs.berkeley.edu/dist/vint/ns-src-2.1b6.tar.gz. Nader >>> "Jiangbin" == Jiangbin Yang writes: Jiangbin> Hello, The files (e.g., the download file for ns-2.1b6) in Jiangbin> http://www-mash.CS.Berkeley.EDU/ns/CHANGES.html could not be Jiangbin> downloaded. Below is the error msg returned: >> Not Found >> The requested URL /dist/ns-src-2.1b6.tar.gz was not found on this >> server. >> Apache/1.3.9 Server at www-mash.CS.Berkeley.EDU Port 80 Jiangbin> Could you please check this? Jiangbin> Thanks, Jiangbin> Jiangbin Yang ______________________ Broadband Networks Lab Jiangbin> Dept. of Systems & Computer Engineering Carleton University Jiangbin> Ottawa, Ontario K1S 5B6, Canada Tel: (613) 520-2600 x2913 Jiangbin> Fax: (613) 520-5727 http://www.sce.carleton.ca/~yang Jiangbin> ________________________________ From barenco@dit.upm.es Tue Jan 25 22:24:00 2000 From: barenco@dit.upm.es (Claudia Jacy Barenco) Date: Tue, 25 Jan 2000 23:24:00 +0100 Subject: Multifield Classifier Message-ID: <001001bf6782$e1bc80e0$ead0523e@fofinha> Dear users, Is there any implementation of a Multifield classifier (src_adrr, dst_addr, DSCP bits) ? Thanks for attention. ***************************************** Cl醬dia Barenco Polythecnic University of Madrid E.T.S.I. Telecommunication Telematic Systems Madrid - Spain ***************************************** From Fusun Inanc Tue Jan 25 22:35:28 2000 From: Fusun Inanc (Fusun Inanc) Date: Tue, 25 Jan 2000 17:35:28 -0500 (EST) Subject: nam usage Message-ID: <200001252235.RAA11837@poptop.> Hi everyone, I have another question. In my Sun Solaris 2.7 unix system there are three users using ns2. I have just realised that if one user uses nam in his/her simulation first and if the other user wants to use it at that time, the second user's nam window pops up on the first user's screen. I copied the nam executable files in different directories for each user but the problem still exist. I think if nam is running as a process on the unix server, it cannot run another process.. Is that true? Or is there something that I can do? Thanks for your kind help. Fusun Inanc From Tarik Alj Tue Jan 25 23:05:00 2000 From: Tarik Alj (Tarik Alj) Date: Tue, 25 Jan 2000 18:05:00 -0500 (EST) Subject: Multifield Classifier Message-ID: <200001252305.SAA09996@someware.INRS-Telecom.UQuebec.CA> See hash-classifier: src, dst, fid > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > From: "Claudia Jacy Barenco" > To: > Subject: Multifield Classifier > Date: Tue, 25 Jan 2000 23:24:00 +0100 > MIME-Version: 1.0 > Content-Transfer-Encoding: 8bit > X-Priority: 3 > X-MSMail-Priority: Normal > X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 > > Dear users, > > Is there any implementation of a > Multifield classifier (src_adrr, dst_addr, DSCP bits) ? > Thanks for attention. > > ***************************************** > Cl醬dia Barenco > Polythecnic University of Madrid > E.T.S.I. Telecommunication > Telematic Systems > Madrid - Spain > ***************************************** Tarik From litmaath@fnal.gov Tue Jan 25 23:26:45 2000 From: litmaath@fnal.gov (Maarten Litmaath) Date: Tue, 25 Jan 2000 17:26:45 -0600 Subject: 512 parallel FTP sessions very slow? References: <3884E956.90835C76@fnal.gov> Message-ID: <388E3135.335C0396@fnal.gov> Dear NS experts, recently I asked you for clues that would explain the slowness of the following supposedly simple simulation of 512 parallel FTP sessions (to estimate the maximum performance for a switch simulation we have in mind): #---------------------------------------------------------------------------- set ns [new Simulator] $ns set-address-format expanded set maxnode 512 #set maxnode 511 set f [open out.tr w] $ns trace-all $f set nf [open out.nam w] $ns namtrace-all $nf for {set i 0} {$i < $maxnode} {incr i} { set node($i) [$ns node] set k [expr $i + $maxnode] set node($k) [$ns node] $ns duplex-link $node($i) $node($k) 10Mb 10ms DropTail $ns queue-limit $node($i) $node($k) 100 set tcp [$ns create-connection TCP $node($i) TCPSink $node($k) 0] set ftp [$tcp attach-source FTP] $ns at 0.000 "puts \"\[$ns now] start $i\"; $ftp start" $ns at 0.100 "$ftp stop; puts \"\[$ns now] stop $i\"" puts "node $i defined" } $ns at 0.200 "close $f; close $nf; $ns halt" $ns run #---------------------------------------------------------------------------- Compared to the original code we made a few modifications that do not change the behavior: we tried using 10Mb and 10ms for the duplex link, we set the queue-limit to 100, and we added logging to see what is going on. We find that this code takes 1.5 hours on a Sparc-20, which is much too much and inconsistent with other people being able to simulate thousands of nodes at a time. Guillermo Rodriguez Garcia suggested that a good fraction of this time was due to the topology generation, but according to Mark Allman NS is spending "the vast majority of its time running the simulation, not setting it up" (he dumped debug info periodically). Experimenting further I found that up to 511 (!) nodes the simulation seems to take a reasonable time: 38 minutes for 0.1 (!) second of simulated time. When I add one more node it becomes 88 minutes! Somehow NS behaves differently when the number of node address bits goes from 9 (511) to 10 (512): does that ring a bell? I also noticed that during this "simple" simulation NS grows to 76 MB, and that the logfiles are NOT being written as the simulation proceeds, but only at the end! What kind of buffering is going on here?! The logfile sizes are 5 and 11 MB, respectively. Thanks for any further clues. I will inform you of any progress. Greetings, Maarten From guol@cs.bu.edu Tue Jan 25 23:39:17 2000 From: guol@cs.bu.edu (Guo, Liang) Date: Tue, 25 Jan 2000 18:39:17 -0500 (EST) Subject: question about TCPAgent reset() Message-ID: I have a question on resetting TCPAgent and reusing it. the question is: how to delete those old acknowledgement packets that still in flight in the network? This is important when we need to recycle old TCPAgents immediately after the connection is closed. (like in webtraf.*). Consider the following case: At time 1, connection 1 is set up, it transmits 100 packets, then close the session at time 10. Suppose the round trip time is 1 second. And the last ACK packet (ACK for packet 100) was retransmitted several times due to packet loss or extremely high queueing delay. Then, at time 10.1, we want to set up a new connection, but we want to reuse the TCPAgent that we just used, so the following code is called (as in webtraf.cc/tcl): ... stcp = picktcp(); // load a recycled TCPAgent from some pool ssnk = picksnk(); // load a recycled TCPSink ... $stcp reset $ssnk reset $ns attach-agent $clnt $stcp $ns attach-agent $svr $ssnk $ns connect $stcp $ssnk ... $stcp advanceby XXX; // XXX is the number of pkts to xmit, say 90 ... However, since the ACK packet (acking packet 100) for the previous connection is still in flight, so as soon as $stcp starts, it will get this ACK packet and end the session, but it hasn't sent any packet out! I don't know how to solve this problem, increasing the pool size can alleviate the problem, but can not solve it. Any suggestions? Guo, Liang From swtan@mmu.edu.my Wed Jan 26 09:11:42 2000 From: swtan@mmu.edu.my (Tan Su Wei) Date: Wed, 26 Jan 2000 09:11:42 +0000 Subject: How to manual configure 2 separate paths? References: <388DAAA5.E7840A52@mmu.edu.my> <004d01bf677d$d4e4ca80$973e8489@ericwong> Message-ID: <388EBA4D.466D1603@mmu.edu.my> Hi, My understanding of ManualRtNode is: It is based on Hash/Dest classifier... thus when you first set path one (n1-n2) using add-route-to-adj-node ... , the neigbour node will become the default route to the current node calling this procedure... e.g: -> $r1 add-...-node -default $r2, then a call to -> $r1 add-...-node -default $r3, will reset the hash entry. Thus route 1 and route 2 will be aggregated together.... and if you try to reverse on of the path, you will get result like the traffic flowing back to the source... This all because the hash entry are based on the Destination only (classifier: Hash/Dest).... > > $node add-routes > > esp. the argument. > Take a look at ManualRtNode code in ~/ns-2.1xx/tcl/lib/ns-node.tcl Hope I answer your questions... Regards Tan Su Wei From guille@iies.es Wed Jan 26 02:02:33 2000 From: guille@iies.es (Guillermo Rodriguez Garcia) Date: Wed, 26 Jan 2000 03:02:33 +0100 Subject: 512 parallel FTP sessions very slow? In-Reply-To: <388E3135.335C0396@fnal.gov> References: <3884E956.90835C76@fnal.gov> Message-ID: <3.0.3.32.20000126030233.00822a90@mailhost.iies.es> At 17:26 25/01/00 -0600, you wrote: > >Guillermo Rodriguez Garcia suggested that a good fraction >of this time was due to the topology generation, but according to Mark Allman > NS is spending "the vast majority of its time running >the simulation, not setting it up" (he dumped debug info periodically). I have been extending the scenario generators, so I have often used ns to generate topologies of several hundreds and sometimes one or two thousand nodes *without actually running any simulation*, and then I have also run some expensive simulations. I can say that the topology generation takes a considerable ammount of time. There are ways to check this. For example, just generate the topology and view it with nam, without running the simulation (see the topoview script in the scen-gen package) - you will see what I mean. Regards, G. -- Guillermo Rodriguez Garcia From n2688420@dec4000.cc.ncku.edu.tw Wed Jan 26 04:13:45 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Wed, 26 Jan 2000 12:13:45 +0800 Subject: update problem!! Message-ID: <388E7479.7C849E5F@ccmail.ncku.edu.tw> Hi~ genetlemen I have just updated my tclcl to tclcl-1.0b9-current successfully. but while re-compiling ns-2.1b6-current on my linux 6.0 the following error message display. have anyone met this situation? hope someone could help me and give me a hint to solve this strange problem. thanks for advise in advance regards ****************************************************************** rng.o(.rodata+0x138): undefined reference to `TclObject::delay_bind_dispatch(cha r const *, char const *)' ranvar.o(.rodata+0x1b8): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x1f8): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x238): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)'rng.o(.rodata+0x138): undefined reference to `TclObject::delay_bind_dispatch(cha r const *, char const *)' ranvar.o(.rodata+0x1b8): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x1f8): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x238): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x278): undefined reference to `TclObject::delay_bind_dispatch( char const *, char const *)' ranvar.o(.rodata+0x2b8): more undefined references to `TclObject::delay_bind_dis patch(char const *, char const *)' follow collect2: ld returned 1 exit status make: *** [ns] Error 1 From haoboy@isi.edu Wed Jan 26 04:16:05 2000 From: haoboy@isi.edu (Haobo Yu) Date: Tue, 25 Jan 2000 20:16:05 -0800 (PST) Subject: update problem!! In-Reply-To: <388E7479.7C849E5F@ccmail.ncku.edu.tw> Message-ID: Did you make clean in your ns directory? If it's a mismatch of your ns and tclcl versions you should see error messages in the compiling time but not linking. - Haobo On Wed, 26 Jan 2000, CWen56 wrote: > Hi~ genetlemen > I have just updated my tclcl to tclcl-1.0b9-current successfully. > but while re-compiling ns-2.1b6-current on my linux 6.0 the following > error message display. > have anyone met this situation? hope someone could help me and give > me a hint to solve this strange problem. > thanks for advise in advance > regards > ****************************************************************** > rng.o(.rodata+0x138): undefined reference to > `TclObject::delay_bind_dispatch(cha > r const *, char const *)' > ranvar.o(.rodata+0x1b8): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x1f8): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x238): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)'rng.o(.rodata+0x138): undefined reference to > `TclObject::delay_bind_dispatch(cha > r const *, char const *)' > ranvar.o(.rodata+0x1b8): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x1f8): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x238): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x278): undefined reference to > `TclObject::delay_bind_dispatch( > char const *, char const *)' > ranvar.o(.rodata+0x2b8): more undefined references to > `TclObject::delay_bind_dis > patch(char const *, char const *)' follow > collect2: ld returned 1 exit status > make: *** [ns] Error 1 > > From leventi@fgan.de Wed Jan 26 08:39:39 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Wed, 26 Jan 2000 09:39:39 +0100 (MET) Subject: Bad event type MouseWheel-nam1.08a on Solaris 6 In-Reply-To: <200001250221.CAA25617@churchward.erg.abdn.ac.uk> (message from Martin Koyabe on Tue, 25 Jan 2000 02:21:56 +0000 (GMT)) Message-ID: <200001260839.JAA17764@melle.ffm.fgan.de> Hello dear Martin, hello dear all I have tried to make the changes that you proposed in the tcl file. It has sounded also reasonable to me. I have commended out two lines and added in the test-rlm.tcl file: #$ns trace-all [open out.tr w] #$ns namtrace-all [open out.nam w] set nf [open out.nam w] $ns namtrace-all $nf set scn [new Scenario$scenario $ns] #$ns at [expr $runtime +1] "$ns finish" $ns at 3.0 "finish" proc finish {} { global ns nf $ns flush-trace close $f close $nf puts "running nam..." exec nam out.nam & exit 0 } $ns run Unfortunately the problem is persistent. When I give >ns test-rlm.tcl I get } : bad event type or keysym "MouseWheel" while executing "bind Listbox { %W yview scroll [expr - (%D / 120) * 4] units } and the simulator doesn't start BUT IF I give in the test-rlm script instead of my new nam (=nam-1.08) only the nam_old (=the version for Solaris from www-mash.cs.berkeley.edu/nam/) then it works. Are there some Libraries missing on my Sun Workstation? From huang@tik.ee.ethz.ch Wed Jan 26 09:58:14 2000 From: huang@tik.ee.ethz.ch (Polly Huang) Date: Wed, 26 Jan 2000 10:58:14 +0100 (CET) Subject: question about TCPAgent reset() In-Reply-To: Message-ID: Guo, Liang, I used to play this trick with the large-scale traffic simulations. This is one of the better solutions to distinguish pkts for different 'conceptual' tcp connections from the same 'physical' tcp agent object. 1. Assign unique tcp agent fid_ (or flowid_?) for each 'conceptual' connections. 2. Make TCP sinks to carry the fid_ over at the acknowledgement packets. 3. At recv() of TCP sender, check fid_ of the arriving packets. Discard them if the fid_'s don't match. hope this helps, -Polly On Tue, 25 Jan 2000, Guo, Liang wrote: > > I have a question on resetting TCPAgent and reusing it. > the question is: how to delete those old acknowledgement packets > that still in flight in the network? > > This is important when we need to recycle old TCPAgents immediately > after the connection is closed. (like in webtraf.*). Consider the > following case: At time 1, connection 1 is set up, it transmits > 100 packets, then close the session at time 10. Suppose the round > trip time is 1 second. And the last ACK packet (ACK for packet 100) > was retransmitted several times due to packet loss or extremely high > queueing delay. Then, at time 10.1, we want to set up a new > connection, but we want to reuse the TCPAgent that we just used, > so the following code is called (as in webtraf.cc/tcl): > > ... > stcp = picktcp(); // load a recycled TCPAgent from some pool > ssnk = picksnk(); // load a recycled TCPSink > ... > $stcp reset > $ssnk reset > > $ns attach-agent $clnt $stcp > $ns attach-agent $svr $ssnk > $ns connect $stcp $ssnk > ... > > $stcp advanceby XXX; // XXX is the number of pkts to xmit, say 90 > ... > > However, since the ACK packet (acking packet 100) for the previous > connection is still in flight, so as soon as $stcp starts, it > will get this ACK packet and end the session, but it hasn't > sent any packet out! > > I don't know how to solve this problem, increasing the pool > size can alleviate the problem, but can not solve it. Any > suggestions? > > Guo, Liang > From pippes@mailcity.com Wed Jan 26 10:09:59 2000 From: pippes@mailcity.com (James Pipes) Date: Wed, 26 Jan 2000 10:09:59 0000 Subject: test-suite-cbq.tcl Message-ID: Hi , I am trying to find the file "test-suite-cbq.tcl" but I cannot find it anywhere. Does anyone know where I can get it from? LYCOShop. Thousands of products! One location! http://shop.lycos.com/ From czermin@dune.LKN.E-Technik.TU-Muenchen.DE Wed Jan 26 10:05:24 2000 From: czermin@dune.LKN.E-Technik.TU-Muenchen.DE (Czermin Martin) Date: Wed, 26 Jan 2000 11:05:24 +0100 Subject: Difference between WFQ and DRR or WRR Message-ID: <388EC6E4.5F8701AD@lkn.ei.tum.de> Hi all, I have a Q about the different scheduling algorithm implemented in ns. I have read in the mailing list that there allready exist code for an additional implementation of WFQ. WFQ, as a work conserving algoritm, uses the FFQ model to calculate which packet has to be served in the next step. But I don磘 find any literature, explaining the difference to the other models like DRR or WRR. So can anybody give me a explanation what are the main differences and how the other models work to calculate the processing of the incoming packets. Furthermore can I use DRR or WRR instead WFQ in a ns simulation without getting great differences in measuring the delay of the packets, transfering some queues using DRR or WRR instead WFQ algorithm. Thanks in advance Martin From L.Wood@eim.surrey.ac.uk Wed Jan 26 12:16:24 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 26 Jan 2000 12:16:24 +0000 (GMT) Subject: question about TCPAgent reset() In-Reply-To: Message-ID: On Tue, 25 Jan 2000, Guo, Liang wrote: > However, since the ACK packet (acking packet 100) for the previous > connection is still in flight, so as soon as $stcp starts, it > will get this ACK packet and end the session, but it hasn't > sent any packet out! Surely the sequence number for that ack packet should (in most cases) be beyond the right window of the new $stcp session, so $stcp should simply ignore it? Looking in: void TcpAgent::newack(Packet* pkt) [..] hdr_tcp *tcph = hdr_tcp::access(pkt); [..] last_ack_ = tcph->seqno(); prev_highest_ack_ = highest_ack_ ; highest_ack_ = last_ack_; if (t_seqno_ < last_ack_ + 1) t_seqno_ = last_ack_ + 1; some sort of (untested conceptual suggestion; off the top of my head from a quick browse) sanity check would be sensible here: last_ack_ = tcph->seqno(); if (last_ack_ > maxseq_) break; /* ack isn't acking anything we sent */ /* probably other ways to do it: Got an int - TracedInt > TracedDouble here. ugh. if (last_ack_ - highest_ack_ > cwnd_) break; */ prev_highest_ack_ = highest_ack_ ; highest_ack_ = last_ack_; etc. This isn't quite Savage-et-al style defensiveness; just trying to turn: /* * Process a packet that acks previously unacknowleged data. */ into /* * Process a packet that acks previously unacknowleged data * that we think we sent. */ cheers, L. thinks tcp really should distinguish pkts for different 'conceptual' tcp connections from the same 'physical' tcp agent object anyway. PGP From L.Wood@eim.surrey.ac.uk Wed Jan 26 12:27:41 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 26 Jan 2000 12:27:41 +0000 (GMT) Subject: test-suite-cbq.tcl In-Reply-To: Message-ID: On Wed, 26 Jan 2000, James Pipes wrote: > I am trying to find the file "test-suite-cbq.tcl" but I cannot > find it anywhere. Does anyone know where I can get it from? ~ns/tcl/test do: man find L. PGP From hong" Hi, all I want to install ns-2.1b6 on sparc-sun-solaris2.7. When I make it, it reports the following error. ---- ld: fatal: symbol `random' is multiply defined: (file random.o and file /usr/X11R6/lib/libXbsd.a(Berklib.o)); ld: fatal: File processing errors. No output written to ns collect2: ld returned 1 exit status make: *** [ns] Error 1 ---- Then I tried to omit the function 'random' in random.cc.(I do not know whether it is right) and 'make' again. There are still other errors as follows. ----- Undefined first referenced symbol in file Scheduler::instance_ timer-handler.o Scheduler::schedule(Handler *, Event *, double)timer-handler.o ld: fatal: Symbol referencing errors. No output written to ns collect2: ld returned 1 exit status make: *** [ns] Error 1 ------ Thanks for any help in advance. Hong From L.Wood@eim.surrey.ac.uk Wed Jan 26 16:55:52 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 26 Jan 2000 16:55:52 +0000 (GMT) Subject: question about TCPAgent reset() In-Reply-To: Message-ID: On Wed, 26 Jan 2000, Lloyd Wood wrote: > On Tue, 25 Jan 2000, Guo, Liang wrote: > > > However, since the ACK packet (acking packet 100) for the previous > > connection is still in flight, so as soon as $stcp starts, it > > will get this ACK packet and end the session, but it hasn't > > sent any packet out! > > Surely the sequence number for that ack packet should (in most cases) > be beyond the right window of the new $stcp session, so $stcp should ^^ edge of the there. that makes more sense. L. > simply ignore it? > > Looking in: > > void TcpAgent::newack(Packet* pkt) > [..] > hdr_tcp *tcph = hdr_tcp::access(pkt); > [..] > last_ack_ = tcph->seqno(); > prev_highest_ack_ = highest_ack_ ; > highest_ack_ = last_ack_; > > if (t_seqno_ < last_ack_ + 1) > t_seqno_ = last_ack_ + 1; > > some sort of (untested conceptual suggestion; off the top of my head > from a quick browse) sanity check would be sensible here: > > last_ack_ = tcph->seqno(); > if (last_ack_ > maxseq_) > break; /* ack isn't acking anything we sent */ > /* probably other ways to do it: > Got an int - TracedInt > TracedDouble here. ugh. > if (last_ack_ - highest_ack_ > cwnd_) > break; > */ > prev_highest_ack_ = highest_ack_ ; > highest_ack_ = last_ack_; > > etc. This isn't quite Savage-et-al style defensiveness; just trying to > turn: > > /* > * Process a packet that acks previously unacknowleged data. > */ > > into > > /* > * Process a packet that acks previously unacknowleged data > * that we think we sent. > */ > > cheers, > > L. > > thinks tcp really should distinguish pkts for different 'conceptual' > tcp connections from the same 'physical' tcp agent object anyway. > > PGP > > > > > > PGP From guol@cs.bu.edu Wed Jan 26 17:05:31 2000 From: guol@cs.bu.edu (Guo, Liang) Date: Wed, 26 Jan 2000 12:05:31 -0500 (EST) Subject: question about TCPAgent reset() In-Reply-To: Message-ID: > > Surely the sequence number for that ack packet should (in most cases) > > be beyond the right window of the new $stcp session, so $stcp should > ^^ edge of the > > there. that makes more sense. > > L. > Sure, thanks. But, what if this new TCP session is only going to send 1 packet? (And the old session happens to end up with retransmitting a duplicate ACK for packet number 0). I think this problem will still exists or am I making a mistake here? Maybe Polly's suggestion is better (double checking the flow id). Regards. Liang. From haoboy@isi.edu Wed Jan 26 18:07:38 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 26 Jan 2000 10:07:38 -0800 (PST) Subject: Bad event type MouseWheel-nam1.08a on Solaris 6 In-Reply-To: <200001260839.JAA17764@melle.ffm.fgan.de> Message-ID: I've seen this a couple of times. Perhaps you have another older tcl (earlier than 8.0p2) installation under some default location, e.g., /usr/local/lib? Check the Makefile of you tclcl and make sure the 8.0.5 versions are there and before /usr/local/lib in LIB and INCLUDES. Rebuild tclcl and ns-2. - Haobo On Wed, 26 Jan 2000, Anastasia Leventi-Peetz wrote: > > > Hello dear Martin, hello dear all > > I have tried to make the changes that you proposed in the tcl file. > It has sounded also reasonable to me. > I have commended out two lines and added in the test-rlm.tcl file: > > #$ns trace-all [open out.tr w] > #$ns namtrace-all [open out.nam w] > set nf [open out.nam w] > $ns namtrace-all $nf > > set scn [new Scenario$scenario $ns] > #$ns at [expr $runtime +1] "$ns finish" > $ns at 3.0 "finish" > proc finish {} { > global ns nf > $ns flush-trace > close $f > close $nf > > puts "running nam..." > exec nam out.nam & > exit 0 > } > > $ns run > > Unfortunately the problem is persistent. > When I give > >ns test-rlm.tcl > I get > } > : bad event type or keysym "MouseWheel" > while executing > "bind Listbox { > %W yview scroll [expr - (%D / 120) * 4] units > } > and the simulator doesn't start > BUT IF I give in the test-rlm script instead of my new nam (=nam-1.08) > only the nam_old (=the version for Solaris from www-mash.cs.berkeley.edu/nam/) > then it works. > Are there some Libraries missing on my Sun Workstation? > From L.Wood@eim.surrey.ac.uk Wed Jan 26 18:57:01 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Wed, 26 Jan 2000 18:57:01 +0000 (GMT) Subject: question about TCPAgent reset() In-Reply-To: Message-ID: On Wed, 26 Jan 2000, Guo, Liang wrote: > > > Surely the sequence number for that ack packet should (in most cases) > > > be beyond the right window of the new $stcp session, so $stcp should > > ^^ edge of the > > > > there. that makes more sense. [suggested ack-before-right-edge-of-window check] > But, what if this new TCP session is only going to send 1 packet? (And > the old session happens to end up with retransmitting a duplicate > ACK for packet number 0). I think this problem will still exists or > am I making a mistake here? Hence 'most cases'. The worrying thing to me is that I was assuming that the TCPs were tracking per-flow state anyway, and that unique flow-ids were assigned as a matter of course... L. PGP From salehi@isi.edu Wed Jan 26 19:40:16 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 26 Jan 2000 11:40:16 -0800 (PST) Subject: the validation test problem for ns2 In-Reply-To: <20000119232814.47896.qmail@hotmail.com> References: <20000119232814.47896.qmail@hotmail.com> Message-ID: <200001261940.LAA13038@joy.isi.edu> Are you using Solaris? There is a byte alignment problem in Solaris which causes the two tests to fail. You can take the current snapshot. Take the current snapshot. If you are not using Solaris, or SunOS, then we need more info. Nader >>> "Shugong" == Shugong Xu writes: Shugong> hi, there, I have installed successfully the ns2 snapshot of Shugong> Dec 7, 1999. However, it outputs the following messages when Shugong> I run the validation test suite: Shugong> validate overall report: some tests failed: Shugong> ./test-all-wireless-lan ./test-all-wireless-lan-newnode to Shugong> re-run a specific test, cd tcl/test; ../../ns Shugong> test-all-TEST-NAME Shugong> /home/johnh/WORKING/VINT/test-ns-2/test_ns: validate Shugong> I noticed that the snapshot after Dec 21 had fixed it from Shugong> the following document: Shugong> http://www.isi.edu/~johnh/WORK/VINT/VALIDATION/199912/ Shugong> Is there anone have some patchs or something else to help me Shugong> fix this problem? Thanks in advance. Shugong> Shugong Shugong> ______________________________________________________ Get Shugong> Your Private, Free Email at http://www.hotmail.com From kts@gigabitwireless.com Wed Jan 26 20:39:28 2000 From: kts@gigabitwireless.com (Tai Suk Kim) Date: Wed, 26 Jan 2000 12:39:28 -0800 Subject: queue-server-like implementation in NS Message-ID: <002b01bf683d$7100e100$1e01a8c0@labgeneral> This is a multi-part message in MIME format. ------=_NextPart_000_0028_01BF67FA.629C0420 Content-Type: text/plain; charset="euc-kr" Content-Transfer-Encoding: quoted-printable Hi, everyone. I am working to implement queue-server-like function with NS. That is to say, some packets are coming into a queue, each packet is served during the time given by a distribution or an input file, and=20 the next packet is served just after the time. The times should be given = packet by packet. I am thinking of one solution. In a delay file used in a link model ; = LinkDelay::recv it is to change txt =3D txtime(p) to the time biven by a distribution or = an input file. But it seems difficult to NS beginner, me. Could you recommend any easier way to do it or let me know=20 the detailed way for my solution? One more question: it is possible to have queueing at input buffers instead of output buffers? Thanks in advance. Tai Suk Kim ------=_NextPart_000_0028_01BF67FA.629C0420 Content-Type: text/html; charset="euc-kr" Content-Transfer-Encoding: quoted-printable
Hi, = everyone.
 
I am working to = implement=20 queue-server-like function with NS.
 
That is to say, some = packets are=20 coming into a queue, each packet
is served during the = time given by a=20 distribution or an input file, and
the next packet is = served just after=20 the time. The times should be given packet by packet.
 
I am thinking of one solution. In a = delay file used=20 in a link model ; LinkDelay::recv
it is to change txt =3D txtime(p) to = the time biven=20 by a distribution or
an input file. But it seems difficult = to NS=20 beginner, me.
 
Could you recommend any easier way to = do it or let=20 me know
the detailed way for my = solution?
 
One more question:
it is possible to = have=20 queueing at input buffers instead of output
buffers?
 
Thanks in advance.
 
Tai Suk = Kim
------=_NextPart_000_0028_01BF67FA.629C0420-- From haoboy@isi.edu Wed Jan 26 20:49:40 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 26 Jan 2000 12:49:40 -0800 (PST) Subject: nam usage In-Reply-To: <200001252235.RAA11837@poptop.> Message-ID: Nam communicates with a "server" on a per-user basis. So as long as those users have their own unique home directories, they should not see this phenomenon. - Haobo On Tue, 25 Jan 2000, Fusun Inanc wrote: > Hi everyone, > > I have another question. In my Sun Solaris 2.7 unix system there are three > users using ns2. > > I have just realised that if one user uses nam in his/her simulation first and > if the other user wants to use it at that time, the second user's nam window > pops up on the first user's screen. > > I copied the nam executable files in different directories for each user but > the problem still exist. > > I think if nam is running as a process on the unix server, it cannot > run another process.. Is that true? Or is there something that I can do? > > Thanks for your kind help. > > Fusun Inanc > > From haoboy@isi.edu Wed Jan 26 21:04:19 2000 From: haoboy@isi.edu (Haobo Yu) Date: Wed, 26 Jan 2000 13:04:19 -0800 (PST) Subject: Sending application data over SRM In-Reply-To: <20000125145758.3954.qmail@hotmail.com> Message-ID: > I want to use NS to send application data using a multicast protocol. I > would like to use SRM. However, looking in the manual, SRM has no provision > for receiving application data. But it then goes on to say that the agent > asumes that any application data is placed in the data portion of the > packet, pointed to by packet->accessdata(). They are different. SRM uses the data portion of a packet to send its own session information, e.g., distance estimations. However, it does not provide the facility to allow transmitting arbitrary data supplied from the "outside". You can hack SRM to do that. Please read the SRM code to see how it put application data in a packet and how to get it out (all the send_*() stuff). - Haobo > Does this mean that I can directly send application data using the above > pointer ? Or do I have to create an application wrapper for the SRM agent ? > > Thanks, > > Mark > ______________________________________________________ > Get Your Private, Free Email at http://www.hotmail.com > > From Jeff_Donahoo@baylor.edu Wed Jan 26 21:29:50 2000 From: Jeff_Donahoo@baylor.edu (Jeff Donahoo) Date: Wed, 26 Jan 2000 15:29:50 -0600 Subject: AIMD test failure for 2.1b6 (all in one) under RH 6.0 Message-ID: I downloaded nsallinone 2.1b6 to install on my RH 6.0 for Intel machine. When I ran validate I got: validate overall report: some tests failed: ./test-all-aimd to re-run a specific test, cd tcl/test; ../../ns test-all-TEST-NAME I tried to run the test individually according to the instructions given at the end of validate, but it didn't work (Perhaps I misunderstood them): ns-2.1b6/tcl/test% ../../ns test-ns-2.1b6/tcl/test% ../../ns test-all-aimd cript following "[ $# -ge 1 ] then f" argument while compiling "if [ $# -ge 1 ]" (file "test-all-aimd" line 40) couldn't read file "test-ns-2.1b6/tcl/test%": no such file or directory I looked at validate and decided to run an individual test I should do the following. This popped up some XGraphs that I closed: ns-2.1b6/tcl/test% ./test-all-aimd Tests: tcp tcpA tcpB tcp_tahoe tcpA_tahoe tcp_reno tcpA_reno tcp_newreno tcpA_ne wreno Running test tcp: ../../ns test-suite-aimd.tcl tcp Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcpA: ../../ns test-suite-aimd.tcl tcpA Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcpB: ../../ns test-suite-aimd.tcl tcpB Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcp_tahoe: ../../ns test-suite-aimd.tcl tcp_tahoe Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcpA_tahoe: ../../ns test-suite-aimd.tcl tcpA_tahoe Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcp_reno: ../../ns test-suite-aimd.tcl tcp_reno Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcpA_reno: ../../ns test-suite-aimd.tcl tcpA_reno Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcp_newreno: ../../ns test-suite-aimd.tcl tcp_newreno Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 Running test tcpA_newreno: ../../ns test-suite-aimd.tcl tcpA_newreno Test output agrees with reference output next? Unknown event type: 16 Unknown event type: 16 All test output agrees with reference output. Suggestions? Thanks, Jeff PS The nsallinone 2.1b6 compiles as is except for the TCL problem documented on the installation page for ENOTSUP which I fixed as instructed: The solution to work around this is to /* #ifdef ENOTSUP case ENOTSUP: return "ENOTSUP"; #endif */ /* #ifdef ENOTSUP case ENOTSUP: return "operation not supported"; #endif */ comment out both of the above statements in tclPosixStr.c. They should be around lines 328 and 775, respectively. PPS Web page nits: 1. On the download page it says for ns-allinone: Download source: current release 2.1b5 (released 18-Jan-2000) but it points to 2.1b6. 2. For the ns-2 from "Getting the pieces" it says: (ns-2.1b6 released 18-Jan-200) ^^^^ - Must be one of those Y2K problems I've heard about. Michael J. Donahoo Baylor University, Dept. of Computer Science Internet: Jeff_Donahoo@Baylor.edu Office: CSI 220M Work: (254) 710-6836 Home: (254) 848-2428 Fax: (254) 710-3889 From salehi@isi.edu Wed Jan 26 21:44:20 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 26 Jan 2000 13:44:20 -0800 (PST) Subject: AIMD test failure for 2.1b6 (all in one) under RH 6.0 In-Reply-To: References: Message-ID: <200001262144.NAA13216@joy.isi.edu> Jeff, Read the bug fixes for ns-2.1b6 (http://www-mash.CS.Berkeley.EDU/ns/ns-problems.html) and see if this will help you. Also thanks for pointing out the typos. I fixed them and just have to wait for the system's nightly update. Nader >>> "Jeff" == Jeff Donahoo writes: Jeff> I downloaded nsallinone 2.1b6 to install on my RH 6.0 for Intel Jeff> machine. When I ran validate I got: Jeff> validate overall report: some tests failed: ./test-all-aimd to Jeff> re-run a specific test, cd tcl/test; ../../ns test-all-TEST-NAME Jeff> I tried to run the test individually according to the Jeff> instructions given at the end of validate, but it didn't work Jeff> (Perhaps I misunderstood them): Jeff> ns-2.1b6/tcl/test% ../../ns test-ns-2.1b6/tcl/test% ../../ns Jeff> test-all-aimd cript following "[ $# -ge 1 ] then f" argument Jeff> while compiling "if [ $# -ge 1 ]" (file "test-all-aimd" line 40) Jeff> couldn't read file "test-ns-2.1b6/tcl/test%": no such file or Jeff> directory Jeff> I looked at validate and decided to run an individual test I Jeff> should do the following. This popped up some XGraphs that I Jeff> closed: Jeff> ns-2.1b6/tcl/test% ./test-all-aimd Tests: Jeff> tcp tcpA tcpB tcp_tahoe tcpA_tahoe tcp_reno tcpA_reno Jeff> tcp_newreno tcpA_ne wreno Jeff> Running test tcp: ../../ns test-suite-aimd.tcl tcp Test output Jeff> agrees with reference output next? Unknown event type: 16 Jeff> Unknown event type: 16 Jeff> Running test tcpA: ../../ns test-suite-aimd.tcl tcpA Test output Jeff> agrees with reference output next? Unknown event type: 16 Jeff> Unknown event type: 16 Jeff> Running test tcpB: ../../ns test-suite-aimd.tcl tcpB Test output Jeff> agrees with reference output next? Unknown event type: 16 Jeff> Unknown event type: 16 Jeff> Running test tcp_tahoe: ../../ns test-suite-aimd.tcl tcp_tahoe Jeff> Test output agrees with reference output next? Unknown event Jeff> type: 16 Unknown event type: 16 Jeff> Running test tcpA_tahoe: ../../ns test-suite-aimd.tcl tcpA_tahoe Jeff> Test output agrees with reference output next? Unknown event Jeff> type: 16 Unknown event type: 16 Jeff> Running test tcp_reno: ../../ns test-suite-aimd.tcl tcp_reno Jeff> Test output agrees with reference output next? Unknown event Jeff> type: 16 Unknown event type: 16 Jeff> Running test tcpA_reno: ../../ns test-suite-aimd.tcl tcpA_reno Jeff> Test output agrees with reference output next? Unknown event Jeff> type: 16 Unknown event type: 16 Jeff> Running test tcp_newreno: ../../ns test-suite-aimd.tcl Jeff> tcp_newreno Test output agrees with reference output next? Jeff> Unknown event type: 16 Unknown event type: 16 Jeff> Running test tcpA_newreno: ../../ns test-suite-aimd.tcl Jeff> tcpA_newreno Test output agrees with reference output next? Jeff> Unknown event type: 16 Unknown event type: 16 Jeff> All test output agrees with reference output. Jeff> Suggestions? Jeff> Thanks, Jeff Jeff> PS The nsallinone 2.1b6 compiles as is except for the TCL Jeff> problem documented on the installation page for ENOTSUP which I Jeff> fixed as instructed: Jeff> The solution to work around this is to Jeff> /* #ifdef ENOTSUP case ENOTSUP: return "ENOTSUP"; #endif */ Jeff> /* #ifdef ENOTSUP case ENOTSUP: return "operation not Jeff> supported"; #endif */ Jeff> comment out both of the above statements in tclPosixStr.c. They Jeff> should be around lines 328 and 775, respectively. INSTALL PAGE> Jeff> PPS Jeff> Web page nits: Jeff> 1. On the download page it says for ns-allinone: Jeff> Download source: current release 2.1b5 (released 18-Jan-2000) Jeff> but it points to 2.1b6. Jeff> 2. For the ns-2 from "Getting the pieces" it says: Jeff> (ns-2.1b6 released 18-Jan-200) ^^^^ - Must be one of those Y2K Jeff> problems I've heard about. Jeff> Michael J. Donahoo Baylor University, Dept. of Computer Science Jeff> Internet: Jeff_Donahoo@Baylor.edu Office: CSI 220M Work: (254) Jeff> 710-6836 Home: (254) 848-2428 Fax: (254) 710-3889 From Jeff_Donahoo@baylor.edu Wed Jan 26 21:48:45 2000 From: Jeff_Donahoo@baylor.edu (Jeff Donahoo) Date: Wed, 26 Jan 2000 15:48:45 -0600 Subject: nam usage In-Reply-To: Message-ID: I have noticed this behavior as well under Linux, and the users do have their own home directories (mounted over NFS). Here's a hack we use: 1. Go to the file anim-ctrl.tcl in the nam-1.0a7/tcl directory. 2. Jump to function remote-create-animator 3. Comment out the whole if/else statement 4. Insert the following before the commented if/else statement: puts -nonewline "Cannot connect to existing nam instance. " puts "Starting a new one..." $self local-create-animator $trace_file [join $args] You may notice that these are the lines in the "if" part of the if/else statement. 5. In directory nam-1.0a7, type make. This should recompile NAM so that every time you run it a NEW NAM CONSOLE is created. This creates a new console each time you run nam, but it avoids the problem of multiple users. Anyone have any better suggestions? Perhaps the problem has been fixed. Jeff > Nam communicates with a "server" on a per-user basis. So as long as those > users have their own unique home directories, they should not see this > phenomenon. > > - Haobo > > On Tue, 25 Jan 2000, Fusun Inanc wrote: > > > Hi everyone, > > > > I have another question. In my Sun Solaris 2.7 unix system > there are three > > users using ns2. > > > > I have just realised that if one user uses nam in his/her > simulation first and > > if the other user wants to use it at that time, the second > user's nam window > > pops up on the first user's screen. > > > > I copied the nam executable files in different directories for > each user but > > the problem still exist. > > > > I think if nam is running as a process on the unix server, it cannot > > run another process.. Is that true? Or is there something that I can do? > > > > Thanks for your kind help. > > > > Fusun Inanc > > > > > > From Fusun Inanc Wed Jan 26 22:17:48 2000 From: Fusun Inanc (Fusun Inanc) Date: Wed, 26 Jan 2000 17:17:48 -0500 (EST) Subject: nam usage Message-ID: <200001262217.RAA12820@poptop.> Thanks a lot Jeff. This has solved my problem.. Now we can open nam files in different screens. Fusun > From: "Jeff Donahoo" > To: "Haobo Yu" , "Fusun Inanc" > Cc: > Subject: RE: nam usage > Date: Wed, 26 Jan 2000 15:48:45 -0600 > MIME-Version: 1.0 > Content-Transfer-Encoding: 7bit > X-Priority: 3 (Normal) > X-MSMail-Priority: Normal > X-MIMEOLE: Produced By Microsoft MimeOLE V5.00.2314.1300 > Importance: Normal > > I have noticed this behavior as well under Linux, and the users do have > their own home directories (mounted over NFS). > > Here's a hack we use: > > > 1. Go to the file anim-ctrl.tcl in the nam-1.0a7/tcl directory. > 2. Jump to function remote-create-animator > 3. Comment out the whole if/else statement > 4. Insert the following before the commented if/else statement: > > puts -nonewline "Cannot connect to existing nam instance. " > puts "Starting a new one..." > $self local-create-animator $trace_file [join $args] > > You may notice that these are the lines in the "if" part of the if/else > statement. > > 5. In directory nam-1.0a7, type make. This should recompile NAM so that > every time > you run it a NEW NAM CONSOLE is created. > > This creates a new console each time you run nam, but it avoids the problem > of multiple users. Anyone have any better suggestions? Perhaps the problem > has been fixed. > > Jeff > > > Nam communicates with a "server" on a per-user basis. So as long as those > > users have their own unique home directories, they should not see this > > phenomenon. > > > > - Haobo > > > > On Tue, 25 Jan 2000, Fusun Inanc wrote: > > > > > Hi everyone, > > > > > > I have another question. In my Sun Solaris 2.7 unix system > > there are three > > > users using ns2. > > > > > > I have just realised that if one user uses nam in his/her > > simulation first and > > > if the other user wants to use it at that time, the second > > user's nam window > > > pops up on the first user's screen. > > > > > > I copied the nam executable files in different directories for > > each user but > > > the problem still exist. > > > > > > I think if nam is running as a process on the unix server, it cannot > > > run another process.. Is that true? Or is there something that I can do? > > > > > > Thanks for your kind help. > > > > > > Fusun Inanc > > > > > > > > > > > From maria@cs.columbia.edu Wed Jan 26 23:45:32 2000 From: maria@cs.columbia.edu (Maria Papadopouli) Date: Wed, 26 Jan 2000 18:45:32 -0500 (EST) Subject: starting the mobile nodes at any given time Message-ID: <200001262345.SAA14628@bourbon.cs.columbia.edu> Hello to all: I was wondering if you have any suggestion on the following problem: I am simulating a protocol related to the mobile/wireless networks using ns-2/cmu-extension and each mobile node starts participating in the protocol (broadcasting pkts/respond to msgs, etc) at a specified time $t$ (the time may be different for each node). However, i would like to force the mobile node not to receive any pkts, if it is not participating in the protocol. I thought that a natural approach is the following: # configure node $ns_ node-config -adhocRouting $opt(rp)\ -llType $opt(ll) \ -macType $opt(mac) \ -ifqType $opt(ifq) \ -ifqLen $opt(ifqlen) \ -antType $opt(ant) \ -propType $prop \ -phyType $opt(netif) \ -channelType $channel \ -topoInstance $topo \ -agentTrace ON \ -routerTrace ON \ -macTrace ON \ -movementTrace ON $ns_ at 5.0 "set node_(1) [$ns_ node]" (assuming that at time 5.0 the node_(1) starts participating in the protocol) but this gives me problems in tcl, because when later i have something of the sort: $ns_ at 5.653178417 "$node_(1) set Z_ 0" ns gives me error: > can't read "node_(1)": no such variable > while executing > "$ns_ at 5.653178417 "$node_(1) set Z_ 0"" I mentioned that to the monarch project, at CMU, and they told me that in their version the above works fine. Would this be a bug in the VINT version? As an alternative, i also tried to set initially the rxPower and txPower to 0.0 and set up to a different number later when a node starts participating in the protocol. However, that didn't work either ... The node could still receive the pkts (even that its rxPower was 0). Do you have any idea/suggestion how to solve this problem. Thanks a lot! Maria Papadopouli --- Computer Science Dept. Columbia University New York, NY10027 http://www.cs.columbia.edu/~maria/ From yaxu@isi.edu Thu Jan 27 00:07:05 2000 From: yaxu@isi.edu (Ya Xu) Date: Wed, 26 Jan 2000 16:07:05 -0800 (PST) Subject: starting the mobile nodes at any given time In-Reply-To: <200001262345.SAA14628@bourbon.cs.columbia.edu> Message-ID: > > As an alternative, i also tried to set initially the rxPower and txPower to 0.0 > and set up to a different number later when a node > starts participating in the protocol. However, that didn't work either ... > The node could still receive the pkts (even that its rxPower was 0). > set node initial energy = 0 -Ya From youngbae@cs.tamu.edu Thu Jan 27 00:17:00 2000 From: youngbae@cs.tamu.edu (Youngbae - Ko) Date: Wed, 26 Jan 2000 18:17:00 -0600 (CST) Subject: NS installation problems w/ a FreeBSD machine In-Reply-To: Message-ID: Hello, I tried to install the current version of ns-2.1b6 on a FreeBSD 2.2.7 machine and saw that the executable 'ns' file was successfully generated under the ns-2.1b6 directory. (actually, I tried both by *all-in-one* and by *piece-by-piece* and got the same result as explained below.) However, when I run or files, something is wrong!! Shortly, with running ./ns, all Tcl commands are shown on the screen and finally stop with the below error messages: : : no value given for parameter "text" to "checkout_executable" while executing "checkout_executable PERL perl "\ When configured, ns found the right version of perl in\ but it doesn't seem to be there anymore, so\ ns will fail b..." I have never experienced this kind of installation problem with Solaris. Anyway, if you know what the problem is and how to fix it, please let me know. Thanks, - Youngbae P.S> FYI, here is a part of commands shown on the screen when I run ./ns. proc warn {msg} { global warned_ if {![info exists warned_($msg)]} { puts stderr "warning: $msg" set warned_ ($msg) 1 } } : : : : On Tue, 18 Jan 2000, Ya Xu wrote: > I know tora has problem w/ Solaris. But it works fine w/ Fbsd & linux > (AFAIK). So far I do not know why it does not work w/ Solaris. I do not > have solaris machine to trace down the problem. > > -Ya From xxe101@cs.usask.ca Thu Jan 27 01:34:48 2000 From: xxe101@cs.usask.ca (Mr. Epsilon) Date: Wed, 26 Jan 2000 19:34:48 -0600 (CST) Subject: waveLAN packet size Message-ID: Hi, Does anybody know what is the size of a waveLAN packet (MTU)? Raja From salehi@isi.edu Thu Jan 27 03:11:08 2000 From: salehi@isi.edu (Nader Salehi) Date: Wed, 26 Jan 2000 19:11:08 -0800 (PST) Subject: Segmentation in TCP In-Reply-To: <5F05C89FB2F8D211B6430008C7919127020EE00E@esealnt190> References: <5F05C89FB2F8D211B6430008C7919127020EE00E@esealnt190> Message-ID: <200001270311.TAA13955@joy.isi.edu> Check Two-way TCP. Not sure though. Nader >>> "ERA" == ERA writes: ERA> Dears ns-usres, I wonder if there is any TCP agent that can break ERA> down a large packet size into smaller packet size. It would be ERA> more efficient if the TCP agent didn't add the hole header and ERA> waited for the ack for each packet. I mean in that case it would ERA> like if I decided a smaller packet size. ERA> Any comment would be appreciated. ERA> George From lli@shtdu.edu.cn Thu Jan 27 05:57:37 2000 From: lli@shtdu.edu.cn (Li Lei) Date: Thu, 27 Jan 2000 13:57:37 +0800 Subject: exp_.max_p_inv alternating values (bug?) Message-ID: <002201bf688b$6f52bf60$18741cd2@oemcomputer> This is a multi-part message in MIME format. ------=_NextPart_000_001F_01BF68CE.78205600 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 RGVhciBBQS4sDQoNCkkgd291bGQgdGVsbCB5b3UgSSBmb3VuZCB0aGUgc2FtZSBxdWVzdGlvbiBh cyB5b3UgYW5kIHdoZW4gSSB3YW50IHRvIHByaW50DQptYXhfdGhyZXNob2xkIGFuZCBzb21lIG90 aGVyIHBhcmFtZXRlcnMsIEkgZm91bmQgdGhlbSBhbHRlcm5hdGVzIGJldHdlZW4gdGhlaXIgZGVm YXVsdA0KdmFsdWUgYW5kIHRoZWlyIGFzc2lnbmVkIHZhbHVlIGFsc28uIEkgYXNrZWQgdGhlIHF1 ZXN0aW9uLCBub25lIHJlc3BvbnNlIGhhZCBJIHJlY2VpdmVkLg0KDQpMaUxlaQ0K ------=_NextPart_000_001F_01BF68CE.78205600 Content-Type: text/html; charset="gb2312" Content-Transfer-Encoding: base64 PCFET0NUWVBFIEhUTUwgUFVCTElDICItLy9XM0MvL0RURCBIVE1MIDQuMCBUcmFuc2l0aW9uYWwv L0VOIj4NCjxIVE1MPjxIRUFEPg0KPE1FVEEgY29udGVudD0idGV4dC9odG1sOyBjaGFyc2V0PWdi MjMxMiIgaHR0cC1lcXVpdj1Db250ZW50LVR5cGU+DQo8TUVUQSBjb250ZW50PSJNU0hUTUwgNS4w MC4yMDE0LjIxMCIgbmFtZT1HRU5FUkFUT1I+DQo8U1RZTEU+PC9TVFlMRT4NCjwvSEVBRD4NCjxC T0RZIGJnQ29sb3I9I2ZmZmZmZj4NCjxESVY+PEZPTlQgc2l6ZT0yPkRlYXIgQUEuLDxCUj48QlI+ SSB3b3VsZCB0ZWxsIHlvdSBJIGZvdW5kIHRoZSBzYW1lIHF1ZXN0aW9uIGFzIA0KeW91IGFuZCB3 aGVuIEkgd2FudCB0byBwcmludDxCUj5tYXhfdGhyZXNob2xkIGFuZCBzb21lIG90aGVyIHBhcmFt ZXRlcnMsIEkgZm91bmQgDQp0aGVtIGFsdGVybmF0ZXMgYmV0d2VlbiB0aGVpciBkZWZhdWx0PEJS PnZhbHVlIGFuZCB0aGVpciBhc3NpZ25lZCB2YWx1ZSBhbHNvLiBJIA0KYXNrZWQgdGhlIHF1ZXN0 aW9uLCBub25lIHJlc3BvbnNlIGhhZCBJIA0KcmVjZWl2ZWQuPEJSPjxCUj5MaUxlaTwvRk9OVD48 L0RJVj48L0JPRFk+PC9IVE1MPg0K ------=_NextPart_000_001F_01BF68CE.78205600-- From leventi@fgan.de Thu Jan 27 08:18:50 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Thu, 27 Jan 2000 09:18:50 +0100 (MET) Subject: Bad event type MouseWheel-nam1.08a on Solaris 6 In-Reply-To: (message from Haobo Yu on Wed, 26 Jan 2000 10:07:38 -0800 (PST)) Message-ID: <200001270818.JAA18615@melle.ffm.fgan.de> Hallo Haobo, thanks a lot for your hint. I have looked up in the Makefile of tclcl-1.09 directory and could make sure tcl8.0.4 will be used. Is version tcl8.0.5 necessary? And when yes, where did you get it from? I found only the newest ones 8.1 and 8.2 best regards: Anastasia From Gwendal.Le-Grand@lip6.fr Thu Jan 27 09:36:47 2000 From: Gwendal.Le-Grand@lip6.fr (Gwendal LE GRAND) Date: Thu, 27 Jan 2000 10:36:47 +0100 Subject: Connecting LANs using hierarchical addessing Message-ID: Hello I want to simulate the following architecture #Wired domain # HA--------------- FA the home and foreign agents # | lan3 | # Wless lan1 | |wireless lan2 # All the nodes # When i call make-lan, what should the address of lan3 be ? Here we have 2 domains (under HA and FA) with one cluster in each then i do set lan1_ [$ns make-lan ... -address "0.0.0" ] set lan2_ [$ns make-lan ... -address "1.0.0"] But what about lan3 (does it also have a -adress option ?) Thanks in advance GWendal -------Gwendal LE GRAND------- mailto:Gwendal.Le-Grand@lip6.fr tel: +33 (0) 1 44 27 75 12 http://www-rp.lip6.fr/~legrand fax: +33 (0) 1 44 27 74 95 Universite Pierre et Marie Curie, Laboratoire LIP6-CNRS, 8 Rue du Capitaine Scott, 75015 Paris, France From n2688420@dec4000.cc.ncku.edu.tw Thu Jan 27 11:40:13 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Thu, 27 Jan 2000 19:40:13 +0800 Subject: ns-2.1b6-current?? Message-ID: <38902E9D.4A4C21FD@ccmail.ncku.edu.tw> Hi!! Gentlemen!! If i install ns-allinone-2.1b6 on my Redhat 6.0, should i have to also install cmu-extension for my wireless simulation? If it needs, is its installion sequence same as cmu's manual's description ? because i found similar files as cmu's files in ns-2.1b6-current. but after installing and re-run past programs, it has core down. may any share his exerience? thanks regards From Kwang-Mien.Chan.GP@icn.siemens.de Thu Jan 27 13:24:20 2000 From: Kwang-Mien.Chan.GP@icn.siemens.de (Chan Kwang Mien) Date: Thu, 27 Jan 2000 14:24:20 +0100 Subject: Why create a shadow object Message-ID: hi, can anyone tell me what is the purpose of mirroring an object in the compiled hierarchy when an object is created in the interpreted hierarchy ? thank you. rgds, kwang mien From leventi@fgan.de Thu Jan 27 13:42:24 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Thu, 27 Jan 2000 14:42:24 +0100 (MET) Subject: ghost-nodes moving by wireless-scenario? Message-ID: <200001271342.OAA19049@melle.ffm.fgan.de> I wonder if I do something badly wrong. I have tried simple-wireless and wireless examples and in both cases, though I get a nam.out file the animator shows nothing as time goes by Any suggestions? Many thanks:A.Leventi-Peetz From Fabienne Lefevre-AFL005 Thu Jan 27 14:31:48 2000 From: Fabienne Lefevre-AFL005 (Fabienne LEFEVRE) Date: Thu, 27 Jan 2000 15:31:48 +0100 Subject: ghost-nodes moving by wireless-scenario? References: <200001271342.OAA19049@melle.ffm.fgan.de> Message-ID: <389056D4.E86E4AA4@crm.mot.com> Dear Anastasia, Actually I had the same problem as you have. If I remeber well, if you want to see the nodes you should increase their size. Use the command $ns initial_node_pos $node 20 to this purpose and check if changing the value 20 make the nodes visible to you. Secondly, not seeing packets is absolutely normal since nam does not support yet packet visualisation when nodes are mobile. You should only see packets in the fixed part of the network. Hope it helps Fabienne ************************************ Fabienne LEFEVRE Centre de recherche Motorola Paris tel: +33 (0)1 69 35 48 17 e-mail: fabienne.lefevre@crm.mot.com ************************************ Anastasia Leventi-Peetz wrote: > > I wonder if I do something badly wrong. I have tried simple-wireless > and wireless examples and in both cases, though I get a nam.out file > the animator shows nothing as time goes by > Any suggestions? > Many thanks:A.Leventi-Peetz -- From leventi@fgan.de Thu Jan 27 15:07:36 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Thu, 27 Jan 2000 16:07:36 +0100 (MET) Subject: ghost-nodes moving by wireless-scenario? In-Reply-To: <389056D4.E86E4AA4@crm.mot.com> (message from Fabienne LEFEVRE on Thu, 27 Jan 2000 15:31:48 +0100) Message-ID: <200001271507.QAA19126@melle.ffm.fgan.de> Dear Fabienne, very many thanks for your hint! Following your advice I get now something to see with the animator (simple-wireless.tcl- it is the example with the two simple moving nodes) but what I see doesn't make much sense to me! It is only two homocentric circles the one with the bigger radius has turned to green and time (on the frame down left) seems frozen! In Marc Greis' tutorial I read something about an xgraph window that a procedure should open for me (when wireless) but nothing like that has happened. Do you have much experience with wireless scenarios? The examples are very few and are also buggy unfortunately. I've tried wireless-test.tcl and wireless.tcl also. It is not amusing to have to find braces to close up in the code etc. Best regards and thanks again Anastasia From rik.wade@theplanet.net Thu Jan 27 15:42:30 2000 From: rik.wade@theplanet.net (rik.wade@theplanet.net) Date: Thu, 27 Jan 2000 15:42:30 +0000 Subject: [ns-users] Validation Error Message-ID: <20000127154230.A27503@theplanet.net> I have done some work with NS in the past, but have to revisit in order to flesh out some results. However, when I come to build and test NS on the old (and new) Operating Systems (RedHat5.2 and 6.1 respectively), I'm getting the following validation error: ../../ns test-suite-simple.tcl tahoe1 QUIET ns: _o58 recv 0: (_o58 cmd line 1) invoked from within "_o58 cmd recv 0" invoked from within "catch "$self cmd $args" ret" (procedure "_o58" line 2) (SplitObject unknown line 2) invoked from within "_o58 recv 0" ../../ns test-suite-simple.tcl tahoe2 QUIET ns: _o58 recv 0: (_o58 cmd line 1) invoked from within "_o58 cmd recv 0" invoked from within "catch "$self cmd $args" ret" (procedure "_o58" line 2) (SplitObject unknown line 2) invoked from within "_o58 recv 0" [etc etc etc] Could anyone tell me what the problem is and a pointer to where to go to fix it would be great? The "SplitObject" error doesn't lead me anywhere useful yet. thanks, rik wade From Kubinszky Ferenc Thu Jan 27 17:27:49 2000 From: Kubinszky Ferenc (Kubinszky Ferenc) Date: Thu, 27 Jan 2000 18:27:49 +0100 (MET) Subject: Network programming problem Message-ID: <200001271727.SAA06862@bintang> Hi, I have a serious programing problem, and I have no idea... We have a routing daemon. It works. When there isn't any route to the host in the kernel's table, the daemon will find. My problem is that this machine has to be connected to the network simulator (ns2 CMU Extension). In the simulator there are some virtual nodes with the same routing algorithm. The machine and its routing daemon want to talk to these nodes... Routing Daemon | simulator --?-- Kernel <---> User App. (ping vi.rt.no.de) machine ^ I think I should put here something to capture the outgoing packets (HELLOs RREQs, ...)before they leave, and 'wrap' them. (encapsulation?/tunelling? I don't know too much about these techniques) Any idea ? Ferenc Kubinszky ------------- End Forwarded Message ------------- From murphys@teltec.dcu.ie Thu Jan 27 19:05:50 2000 From: murphys@teltec.dcu.ie (Sean Murphy) Date: Thu, 27 Jan 2000 19:05:50 +0000 (BST) Subject: nsaddr_t - why is it an int32_t? Message-ID: Hi, I think nsaddr_t should be a u_int32_t - I've spent most of the afternoon looking for a problem that wouldn't have arisen if nsaddr_t was a u_int32_t - essentially, it was a test for mcast in the rsvp/ns code that worked by comparing the node addr with 1 << McastShift (addr >= 1< Message-ID: On Thu, 27 Jan 2000, Sean Murphy wrote: > I think nsaddr_t should be a u_int32_t - I've spent most of the afternoon > looking for a problem that wouldn't have arisen if nsaddr_t was a > u_int32_t - essentially, it was a test for mcast in the rsvp/ns code that > worked by comparing the node addr with 1 << McastShift (addr >= > 1< ok, but if int32_t's are used, then the result is not the desired result. > It's probably better to use a bitwise and based test which would return > the correct result in any case but I still think it more natural for > nsaddr_t to be u_int32_t's. > > Is there a reason why nsaddr_t's should be int32_t's? Are there situations > in which negative node numbers are used? In a similar vein, I'd like to ask about negative time. clock() returns a double, but simulations start halfway at 0; as far as I can tell you can't start simulations at an arbitrary negative time and run them forwards to execute events scheduled in negative time, by setting the start time with e.g. $ns run from -500.0 (or, more usefully, $ns run from 20 if you want to skip creating some traffic in the first part of your simulation but can't be bothered to comment it out; $ns run would be the same as $ns run from 0. Handy.) Might be useful to have more control over clock precision for some long simulations. [This came up a while back when I was convincing Tom that a bit of satellite geometry code needed finetuning to cover for the possibility, since the possibility was beyond our control...] L. is wondering why the commented-out line in ns-lib.tcl: Simulator instproc run {} { #$self compute-routes PGP From n2688420@dec4000.cc.ncku.edu.tw Fri Jan 28 02:14:33 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Fri, 28 Jan 2000 10:14:33 +0800 Subject: core dumped Message-ID: <3890FB89.9E2876FB@ccmail.ncku.edu.tw> Hi I don't what's happening on my Redhat 6.0. When I install ns-src-2.1b6-curren, nothing could pass the validation. After changing installing ns-allinone-2.1b6, all tests pass. but nomatter what i run simulations like wireless.tcl,wireless-test.tcl and all the successful simulations before, it just core dumped. May someone share his experience? or give me a hint about that. because i didn't find the solution in FAQ. thanks regards *********************************** [root@atmlab03 ex]# ns wireless-test.tcl Segmentation fault (core dumped) From debashis@cet.iitkgp.ernet.in Fri Jan 28 10:01:57 2000 From: debashis@cet.iitkgp.ernet.in (Debashis Biswas) Date: Fri, 28 Jan 2000 15:31:57 +0530 (IST) Subject: vlan simulator Message-ID: my question is ----- can the ns simulator simulate a vlan ? if so, how and what changes are needed ? if not, then is there any simulator which does simulate a vlan ? \\\|/// \\ ~ ~ // ( @ @ ) ************************-oOOo-(_)-oOOo-************************************** Debashis Biswas Computer Science and Engineering ( Final Year) Address : D-220 Azad Hall of Residence, Indian Institute of Technology, Kharagpur, West Bengal, PIN - 721302. *********************************-Oooo.-************************************* .oooO ( ) ( ) ) / \ ( (_/ \_) From rgp5@cornell.edu Fri Jan 28 04:43:20 2000 From: rgp5@cornell.edu (Robin POSS) Date: Thu, 27 Jan 2000 23:43:20 -0500 Subject: project length assessment Message-ID: <3.0.3.32.20000127234320.00715bc4@postoffice2.mail.cornell.edu> Hi everybody, I unexpectedly have to fiddle with some ns code, (aodv.{cc,h,...} in fact so far), and so I have 3 questions: *How long do you think it should take for a guy with 2 months worth of C experience, on a PC running linux redHat 5.1 and with emacs as his unique development tool, to understand aodv.cc and crop it wisely so that it do nothing but a simple hello message based neighbor disovery,assuming that he works full-time on it? *Could you advise me a free C++ development environment, for linux, and tell me where I could download it? *How could I tell to my computer:"Rebuild ns but do it quickly I just changed one line in aodv.cc since last time."? Thanks, -- Robin POSS From leventi@fgan.de Fri Jan 28 07:21:36 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Fri, 28 Jan 2000 08:21:36 +0100 (MET) Subject: core dumped In-Reply-To: <3890FB89.9E2876FB@ccmail.ncku.edu.tw> (message from CWen56 on Fri, 28 Jan 2000 10:14:33 +0800) Message-ID: <200001280721.IAA19575@melle.ffm.fgan.de> Hello, Which version of nam do you have installed? I have made exactly the same experience on my PC under Redhat 6.0 I think nam-1.0a8 has problems. Almost all my wireless examples make core dumped, while with on my SUN the same examples run regards A.Leventi-Peetz From leventi@fgan.de Fri Jan 28 07:36:11 2000 From: leventi@fgan.de (Anastasia Leventi-Peetz) Date: Fri, 28 Jan 2000 08:36:11 +0100 (MET) Subject: bug in nam-1.0a8 (wireless-problems) Message-ID: <200001280736.IAA19580@melle.ffm.fgan.de> Hallo dear ns-users and developers, I'm afraid nam-1.0a8 has a bug!!! On my PC Redhat6.0 the wireless examples run with nam-1.0a7 but create core dumped with the new nam! I see many have the problem, so it must be taken into consideration best regards to all A.Leventi-Peetz From L.Wood@eim.surrey.ac.uk Fri Jan 28 12:07:15 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Fri, 28 Jan 2000 12:07:15 +0000 (GMT) Subject: project length assessment In-Reply-To: <3.0.3.32.20000127234320.00715bc4@postoffice2.mail.cornell.edu> Message-ID: On Thu, 27 Jan 2000, Robin POSS wrote: > I unexpectedly have to fiddle with some ns code, > (aodv.{cc,h,...} in fact so far), and so I have 3 questions: > > *How long do you think it should take for a guy with 2 months worth of C > experience, on a PC running linux redHat 5.1 and with emacs as his unique > development tool, to understand aodv.cc and crop it wisely so that it do > nothing but a simple hello message based neighbor disovery,assuming that > he works full-time on it? That depends on how much your changes touch on other parts of ns, and how well you understand the bits of ns you need to. > *Could you advise me a free C++ development environment, for linux, and > tell me where I could download it? gcc (with g++) should already be installed, and you should have used it to compile ns... > *How could I tell to my computer:"Rebuild ns but do it quickly I just > changed one line in aodv.cc since last time."? make does that anyway, since it's just a question of regenerating one piece of object code and doing the final linking, based on checking file timestamps for changes. (Obgrumble: change a #define constant in a header file and I end up having to rebuild all of ns to make sure it gets changed properly; something of a disadvantage to #define constants. I think a lot of them would be better as bound variables instead.) L. PGP From ana_delgado77@hotmail.com Fri Jan 28 12:07:00 2000 From: ana_delgado77@hotmail.com (Ana Delgado) Date: Fri, 28 Jan 2000 12:07:00 GMT Subject: core dumped Message-ID: <20000128120700.38296.qmail@hotmail.com> Hi, Look in the directory for later mails with the subject core dump, we have had the same problem as you. Cheers, Ana >From: CWen56 >To: ns-users >Subject: core dumped >Date: Fri, 28 Jan 2000 10:14:33 +0800 > >Hi > I don't what's happening on my Redhat 6.0. When I install >ns-src-2.1b6-curren, nothing could pass the validation. After changing >installing ns-allinone-2.1b6, all tests pass. but nomatter what i run >simulations like wireless.tcl,wireless-test.tcl and all the successful >simulations before, it just core dumped. > May someone share his experience? or give me a hint about that. >because i didn't find the solution in FAQ. thanks >regards >*********************************** >[root@atmlab03 ex]# ns wireless-test.tcl >Segmentation fault (core dumped) > ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From Tarik Alj Fri Jan 28 13:49:01 2000 From: Tarik Alj (Tarik Alj) Date: Fri, 28 Jan 2000 08:49:01 -0500 (EST) Subject: vlan simulator Message-ID: <200001281349.IAA15335@someware.INRS-Telecom.UQuebec.CA> Hi, there is currently nothing about VLANs in ns. I have written some code for that. It lets you define a subnet, switches, trunk lines, and VLANs. It's built on the same base as actual LANs in ns, and it does not support features like Spanning Tree or GARP applications. The code still has to be reviewed by my supervisor, let me know if you are interested. If you'd rather do the changes yourself, i can tell you most of the code for LANs can be reused, you would have to add a VLAN header at the C++ level; but the rest can mostly be OTcl. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Fri, 28 Jan 2000 15:31:57 +0530 (IST) > From: Debashis Biswas > To: ns-users@mash.cs.berkeley.edu > Subject: vlan simulator > MIME-Version: 1.0 > > > my question is ----- > > can the ns simulator simulate a vlan ? if so, how and what changes are > needed ? if not, then is there any simulator which does simulate a vlan ? > > > > \\\|/// > \\ ~ ~ // > ( @ @ ) > ************************-oOOo-(_)-oOOo-************************************** > > Debashis Biswas > Computer Science and Engineering ( Final Year) > > Address : > D-220 Azad Hall of Residence, > Indian Institute of Technology, > Kharagpur, > West Bengal, > PIN - 721302. > > > > *********************************-Oooo.-************************************* > .oooO ( ) > ( ) ) / > \ ( (_/ > \_) > > Tarik From Tarik Alj Fri Jan 28 13:59:24 2000 From: Tarik Alj (Tarik Alj) Date: Fri, 28 Jan 2000 08:59:24 -0500 (EST) Subject: Questions about LAN Message-ID: <200001281359.IAA15349@someware.INRS-Telecom.UQuebec.CA> About the 100Mbps CSMA/CD, after I change those parameters, it did make > sense and the performance changed as I thought. but for 10Mbps CSMA/CD, I > didn't see that much collision which affects the throughput and drop > probability. Are you sure I only need to change those parameters to make the > 100Mbps CSMA/CD work? (I just changed the #define in the .h file) > You mean mac-802_3.h? If you really want to make sure for 10, Fast, Gig, check with IEEE 802.3 standards. It really wasn't an exhaustive list that I gave you. > Do you know how to monitor the ifq_? The queue monitors only work for a pair > of nodes or a link, but ifq_ sits on a vitual link which is between a real > node and a virtual node LanNode. When I use the normal method to monitor it, > it gave me error showing that it didn't know this link. Do you have any > idea? > Actually I have never tried to monitor the ifq. But monitor-queue really calls init-monitor from class SimpleLink, so my guess would be that you need a "Vlink instproc init-monitor". See ns-link.tcl From prcyw@gps.leeds.ac.uk Fri Jan 28 14:43:25 2000 From: prcyw@gps.leeds.ac.uk (Y Wang) Date: Fri, 28 Jan 2000 14:43:25 +0000 (GMT) Subject: Ask information on NS-2 Message-ID: Dear Sir/madam, I am a Ph.D student in Warwick Univeristy, UK. My research will be involve in using network simulator for wireless communication. Could you give me more information on Ns-2( I have visited you website)? the price, the way to get it, is it suitable for wireless error corrcetion? Please reply me as soon as possible. Ziran Sun Research student Electrical & Electronic Devision Engineering School The University of Warwick Tel: +44-01203-528392 Address: Room 2 Anchor house 7 Park Road Coventry, Uk CV1 2LE From Martin Koyabe Fri Jan 28 00:57:25 2000 From: Martin Koyabe (Martin Koyabe) Date: Fri, 28 Jan 2000 00:57:25 +0000 (GMT) Subject: Ask information on NS-2 Message-ID: <200001280057.AAA00771@churchward.erg.abdn.ac.uk> Ziran Visit http://www-mash.CS.Berkeley.EDU/ns/ I suppose ns-2.X is available for download for non-commercial use !! -- Martin > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Fri, 28 Jan 2000 14:43:25 +0000 (GMT) > From: Y Wang > To: ns-users@mash.CS.Berkeley.EDU > cc: Ziran Sun > Subject: Ask information on NS-2 > MIME-Version: 1.0 > > Dear Sir/madam, > > I am a Ph.D student in Warwick Univeristy, UK. My research will be involve > in using > network simulator for wireless communication. Could you give me more > information on > Ns-2( I have visited you website)? the price, the way to get it, is it > suitable for wireless > error corrcetion? > > Please reply me as soon as possible. > > Ziran Sun > Research student > Electrical & Electronic Devision > Engineering School > The University of Warwick > > Tel: +44-01203-528392 > Address: Room 2 > Anchor house > 7 Park Road > Coventry, Uk > CV1 2LE > > > > > > > --------------------------------------------------------------------- Martin L.W.D Koyabe e-mail: koyabe@erg.abdn.ac.uk Electronics Research Group Fraser Noble Building King's College, AB24 3UE (Off) +44-01224-272813 Aberdeen, UK (Mob) +44-07881-610825 From hgossain@ececs.uc.edu Fri Jan 28 16:28:20 2000 From: hgossain@ececs.uc.edu (Hrishikesh Gossain) Date: Fri, 28 Jan 2000 11:28:20 -0500 (EST) Subject: bug in nam-1.0a8 (wireless-problems) In-Reply-To: <200001280736.IAA19580@melle.ffm.fgan.de> Message-ID: hi, i am wondering if u are able to run the nam of wired-cum-wireless example of Mac Greis tutorial correctly. (its the animation of both wired and wireless ) please let me know. regards, hrishi ---------------------------------------------------------------------- Hrishikesh Gossain Center For Distributated and Mobile Computing Dept. of ECECS, University of Cincinnati Phone No. : (513)556-3437 (Lab.) email: hgossain@ececs.uc.edu ---------------------------------------------------------------------- On Fri, 28 Jan 2000, Anastasia Leventi-Peetz wrote: > > > Hallo dear ns-users and developers, > > > I'm afraid nam-1.0a8 has a bug!!! > On my PC Redhat6.0 the wireless examples run with nam-1.0a7 > but create core dumped with the new nam! > I see many have the problem, so it must be taken into > consideration > > best regards to all > A.Leventi-Peetz > From gorinsky@cs.utexas.edu Fri Jan 28 16:31:43 2000 From: gorinsky@cs.utexas.edu (Sergey Gorinsky) Date: Fri, 28 Jan 2000 10:31:43 -0600 (CST) Subject: simulations do not run without tracing (SunOS) In-Reply-To: Message-ID: Hi, I am running experiments with FQ links where individual RED queues. When I disable tracing, the following error is reported. How can I avoid this problem (the best what I achieved for speeding up my simulations was writing the trace to /dev/null)? Thank you, Sergey ns: _o24 unknown-flow 513 769 0 0: can't read "drpT_": no such variable while executing "$q drop-target $drpT_" (procedure "_o19" line 12) (FQLink new-flow line 12) invoked from within "$fq_ new-flow $src $dst $fid" (procedure "_o24" line 3) (Classifier/Hash/Fid/FQ unknown-flow line 3) invoked from within "_o24 unknown-flow 513 769 0 0" From n2688420@dec4000.cc.ncku.edu.tw Sat Jan 29 04:12:44 2000 From: n2688420@dec4000.cc.ncku.edu.tw (CWen56) Date: Sat, 29 Jan 2000 12:12:44 +0800 Subject: Wireless Visualization?? Message-ID: <389268BC.3D8BC700@ccmail.ncku.edu.tw> Hi! May someone have install and run ad-hockey successfully ?? When i install Tk800, I got the following error message. I think this question should not be a NS-related question, but i don't know how should i do to solve this problem.. i am not familar with perl and linux :( Hope someone could show me a hint. thanks regards Robert. Ma ****************************************************************** PERL_DL_NONLAZY=1 /usr/bin/perl -Iblib/arch -Iblib/lib -I/usr/lib/perl5/5.00503/ i386-linux -I/usr/lib/perl5/5.00503 -e 'use Test::Harness qw(&runtests $verbose) ; $verbose=0; runtests @ARGV;' t/Require.t t/Trace.t t/X.t t/browseentry.t t/cre ate.t t/fileselect.t t/font.t t/geomgr.t t/mega.t t/mwm.t t/photo.t t/widget.t t /zzHList.t t/zzPhoto.t t/zzScrolled.t t/zzText.t t/zzTixGrid.t t/Require...........Tk::TixGrid is deprecated at t/Require.t line 18 ok t/Trace.............ok t/X.................ok t/browseentry.......ok t/create............/usr/bin/perl: error in loading shared libraries: /usr/share /chinese/xa+cv/wrap.so: undefined symbol: XGetGCValues dubious Test returned status 127 (wstat 32512, 0x7f00) Undefined subroutine &Test::Harness::WCOREDUMP called at /usr/lib/perl5/5.00503/ Test/Harness.pm line 288. make: *** [test_dynamic] Error 127 From tccheng@csis.hku.hk Sat Jan 29 05:46:34 2000 From: tccheng@csis.hku.hk (Cheng Tak Cheung (CE)) Date: Sat, 29 Jan 2000 13:46:34 +0800 (HKT) Subject: example3.tcl in Tutorial doesn't work correctly? Message-ID: I have installed ns-2.1b6 successfully. When I tried to run the example3.tcl from NS tutorial, it could run but I can't see the CBR source working in NAM. There's no data flowing from node 0 to node 3. I have tried this example in ns-2.1b5 without any problem. Does the CBR source work in ns-2.1b6? Joseph Cheng From yaxu@isi.edu Sat Jan 29 06:34:59 2000 From: yaxu@isi.edu (Ya Xu) Date: Fri, 28 Jan 2000 22:34:59 -0800 (PST) Subject: Wireless Visualization?? In-Reply-To: <389268BC.3D8BC700@ccmail.ncku.edu.tw> Message-ID: You also can try nam to visualize wireless simulation -Ya On Sat, 29 Jan 2000, CWen56 wrote: > > Hi! > May someone have install and run ad-hockey successfully ?? When i > install Tk800, I got the following error message. I think this question > should not be a NS-related question, but i don't know how should i do to > solve this problem.. i am not familar with perl and linux :( > Hope someone could show me a hint. > thanks > > regards > Robert. Ma > > ****************************************************************** > > PERL_DL_NONLAZY=1 /usr/bin/perl -Iblib/arch -Iblib/lib > -I/usr/lib/perl5/5.00503/ > i386-linux -I/usr/lib/perl5/5.00503 -e 'use Test::Harness qw(&runtests > $verbose) > ; $verbose=0; runtests @ARGV;' t/Require.t t/Trace.t t/X.t > t/browseentry.t t/cre > ate.t t/fileselect.t t/font.t t/geomgr.t t/mega.t t/mwm.t t/photo.t > t/widget.t t > /zzHList.t t/zzPhoto.t t/zzScrolled.t t/zzText.t t/zzTixGrid.t > t/Require...........Tk::TixGrid is deprecated at t/Require.t line 18 > ok > t/Trace.............ok > t/X.................ok > t/browseentry.......ok > t/create............/usr/bin/perl: error in loading shared libraries: > /usr/share > /chinese/xa+cv/wrap.so: undefined symbol: XGetGCValues > dubious > Test returned status 127 (wstat 32512, 0x7f00) > Undefined subroutine &Test::Harness::WCOREDUMP called at > /usr/lib/perl5/5.00503/ > Test/Harness.pm line 288. > make: *** [test_dynamic] Error 127 > > From aiman-sirhan@usa.net Sat Jan 29 10:07:20 2000 From: aiman-sirhan@usa.net (aiman sirhan) Date: 29 Jan 00 03:07:20 MST Subject: Attention to Ya Xu or Who can help me on NS Message-ID: <20000129100720.10117.qmail@nwcst312.netaddress.usa.net> Dear sir , i have read your totourial about Ns entitled : NS Tutorial : mobile and wireless Network september 10th, 1999. in the page summary (2) it's mentioned that Ns can be used ti simulate Base station node to bridge wired domain and wireless domain. i have downloaded all the documenation about ns , i havent seen any detailed information about how to write a program using a bass station for bridgenig. i have also downloaded the CMU extension to ns documentaion . no mention for base station also. please i need that information. what i need exactly is to write a simple program simulate a node represent the wired domain, another node represent a base station , and finally a third node represent the mobile node so to measure the performance of the TCP/IP for diffrent cases. so plz if you can provide me the help on that. thanks a lot . hoping to hear from you soon. best regards, Eng. aiman sirhan philadelphia universaity ____________________________________________________________________ Get free email and a permanent address at http://www.netaddress.com/?N=1 From puginifabio@tiscalinet.it Sat Jan 29 10:31:31 2000 From: puginifabio@tiscalinet.it (Fabio Pugini) Date: Sat, 29 Jan 2000 11:31:31 +0100 Subject: DLB again Message-ID: <000701bf6a44$2063e040$067c0b3e@default> This is a multi-part message in MIME format. ------=_NextPart_000_0004_01BF6A4C.63E87260 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Hi=20 As I wrote some time ago I'm trying to simulate a Dual Leaky Bucket = (Proposed by A. Elwalid, Bell Laboratories and D. Mitra Lucent = Technologies). It is a 3-parameter traffic-policer (and not a shaper) = but I think it could be well simulated by 2 TBFs in succession with a = particular parameters' choice (i.e. 1st queue=3D2nd queue=3D2nd = buffer=3D1 ). So: How can I put two TBF in succession? (I've read the mailing archive but the proposed solutions don't seem to = be valid). Thanks ------=_NextPart_000_0004_01BF6A4C.63E87260 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
Hi
As I wrote some time ago I'm trying = to simulate=20 a Dual Leaky Bucket (Proposed by A. Elwalid, Bell Laboratories and D. = Mitra=20 Lucent Technologies). It is a 3-parameter traffic-policer (and not a = shaper) but=20 I think it could be well simulated by 2 TBFs in succession with a=20 particular  parameters' choice (i.e. 1st queue=3D2nd queue=3D2nd = buffer=3D1 ).=20 So:
How can I put two TBF in=20 succession?
(I've read the mailing archive but = the proposed=20 solutions don't seem to be valid).
Thanks
------=_NextPart_000_0004_01BF6A4C.63E87260-- From Hajer.Tounsi@ensi.rnu.tn Sat Jan 29 12:28:12 2000 From: Hajer.Tounsi@ensi.rnu.tn (Hajer Tounsi) Date: Sat, 29 Jan 2000 13:28:12 +0100 Subject: classifier? References: <200001211637.LAA06336@someware.INRS-Telecom.UQuebec.CA> <388DCF25.3D2568F2@ensi.rnu.tn> Message-ID: <3892DCDC.7C823E68@ensi.rnu.tn> Hi, could someone tell me how the classifier address is populated? when I create 2 nodes and a duplex link between them, and without adding any route ( by the procedure add-route") how the classifier address knows that the target of the classifier is the link and in which slot it is installed ? I like to do the same thing, but for another default classifier which I have created and which will be the default classifier for the first node. The second node has by default the classifier/Addr and I like to create a duplex link between them, but my problem is that the classifier of the first node (which i create) doesn't found in his slots the target. I don't find how I should populate the classifier which I have created? Thanks for help hajer From Dohy.Hong@ens.fr Sat Jan 29 14:16:31 2000 From: Dohy.Hong@ens.fr (Dohy HONG) Date: Sat, 29 Jan 2000 15:16:31 +0100 (MET) Subject: bug installing ns-2 Message-ID: I have the following warning while doing "make" (for ns-allinone-2.1b4a/ns-2) >In file included from trace.cc:41: >tcp.h:153: ANSI C++ forbids declaration `print_if_needed' with no type >make: *** [trace.o] Error 1 I changed its type to void, then I still have problems with .cc executions: >traffictrace.cc: In method `int TraceFile::setup()': >traffictrace.cc:150: warning: comparison between signed and unsigned >traffictrace.cc:168: implicit declaration of function `int htonl(...)' Could you help me? Thanks a lot, Dohy =================================================== Dohy HONG ECOLE NORMALE SUPERIEURE D閜artement Informatique Mod閘isation et Contr鬺e des R閟eaux 45 rue d'Ulm, 75230 PARIS CEDEX 05, FRANCE Tel : 33 (1) 4432 2112, Fax : 33 (1) 4432 2080 ------------------------------------ Email : Dohy.Hong@ens.fr http://www.di.ens.fr/~mistral/ =================================================== From johnh@isi.edu Sat Jan 29 17:53:58 2000 From: johnh@isi.edu (John Heidemann) Date: Sat, 29 Jan 2000 09:53:58 -0800 Subject: Attention to Ya Xu or Who can help me on NS In-Reply-To: Message from aiman sirhan of "29 Jan 2000 03:07:20 MST." <20000129100720.10117.qmail@nwcst312.netaddress.usa.net> Message-ID: <200001291753.JAA17147@dash.isi.edu> >Dear sir , > >i have read your totourial about Ns entitled : NS Tutorial : mobile and >wireless Network september 10th, 1999. >in the page summary (2) it's mentioned that Ns can be used ti simulate >Base station node to bridge wired domain and wireless domain. >i have downloaded all the documenation about ns , i havent seen any detailed >information about how to write a program using a bass station for bridgenig. >i have also downloaded the CMU extension to ns documentaion . no mention for >base station also. Isn't this exactly what's covered in chapter X of the tutorial? FOr additional examples, grep for "basestation" or "BaseStation" through ~ns-2/tcl/ex and ~ns-2/tcl/test to find a couple. -John Heidemann From barenco@dit.upm.es Sat Jan 29 20:48:18 2000 From: barenco@dit.upm.es (Claudia Jacy Barenco) Date: Sat, 29 Jan 2000 21:48:18 +0100 Subject: Bad Nam trace format Message-ID: <000b01bf6a9a$2cf932a0$16d2523e@fofinha> Dear users, When running Nam I received the following error: "Nam:badly formatted event at offset 1061 in out.nam" The line in out.nam is: "+ 0.288906 0 1 pareto 210 ------- 0 4.0 5.0 0 0" What is wrong with this format ? Thanks in advance. ******************************************** Cl醬dia Jacy Barenco Polytechnic University of Madrid E.T.S.I. Telecommunication Telematic Systems - DIT Madrid - Spain ******************************************** From kts@gigabitwireless.com Sun Jan 30 01:28:01 2000 From: kts@gigabitwireless.com (Tai Suk Kim) Date: Sat, 29 Jan 2000 17:28:01 -0800 Subject: Random txt_time paket by packet in the link Message-ID: <002901bf6ac1$3fe41d90$1e01a8c0@labgeneral> This is a multi-part message in MIME format. ------=_NextPart_000_0026_01BF6A7E.3115A970 Content-Type: text/plain; charset="euc-kr" Content-Transfer-Encoding: quoted-printable Dear NS-users, I need to change the transmission time from a queue.=20 As the first step, i want to change it randomly (with distribution) = packet by packet. Later, i will do it from an input delay file. For the random txt_time, i modified delay.cc and delay.h using = delaymodel.* contributed by Polly Huang. But it makes a core dumped error in double txt =3D ranvar_->value(); Could you let me know what is wrong? Thanks in advance. Tai Suk Kim my modification was as follows. =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D int LinkDelay::command(int argc, const char*const* argv) { Tcl& tcl =3D Tcl::instance(); if (argc =3D=3D 3) { if (strcmp(argv[1], "ranvar") =3D=3D 0) { ranvar_ =3D (RandomVariable*) = TclObject::lookup(argv[2]); return (TCL_OK); } else if (strcmp(argv[1], "bandwidth") =3D=3D 0) { bandwidth_ =3D atof(argv[2]); return (TCL_OK); ------------------------------------------------------------- = =20 } class LinkDelay : public Connector { public: LinkDelay(); void recv(Packet* p, Handler*); void send(Packet* p, Handler*); void handle(Event* e); double delay() { return delay_; } inline double txtime(Packet* p) { return (8. * hdr_cmn::access(p)->size() / bandwidth_); } inline double txtime2(Packet* p) { double txt =3D ranvar_->value(); return txt; //return (8. * hdr_cmn::access(p)->size() / bandwidth_); } double bandwidth() const { return bandwidth_; } void pktintran(int src, int group); protected: int command(int argc, const char*const* argv); RandomVariable* ranvar_; void reset(); double bandwidth_; /* bandwidth of underlying link = (bits/sec) */ double delay_; /* line latency */ Event intr_; int dynamic_; /* indicates whether or not link is ~ */ PacketQueue* itq_; int total_[4]; }; void LinkDelay::recv(Packet* p, Handler* h) { double txt =3D txtime2(p); -------------------------------------------------------------------------= -- =20 } in a tcl file, set dm_ [new DelayLink] set rv [new RandomVariable/Uniform] $rv set min_ 0. $rv set max_ 1. $dm_ ranvar $rv ------=_NextPart_000_0026_01BF6A7E.3115A970 Content-Type: text/html; charset="euc-kr" Content-Transfer-Encoding: quoted-printable
Dear NS-users,
 
I need to change the transmission time = from a=20 queue.
As the first step, i want to change it = randomly=20 (with distribution) packet by packet.
Later, i will do it from an input delay = file.
 
For the random txt_time, i modified = delay.cc and=20 delay.h using delaymodel.* contributed
by Polly Huang. But it makes a core = dumped error=20 in
double txt =3D = ranvar_->value();
 
Could you let me know what is=20 wrong?
 
Thanks in = advance.
 
Tai Suk Kim
 
my modification was as=20 follows.
=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D= =3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D
int = LinkDelay::command(int argc,=20 const char*const* = argv)
{
       =20 Tcl& tcl =3D = Tcl::instance();
        if=20 (argc =3D=3D 3)=20 {
           &n= bsp;   =20 if (strcmp(argv[1], "ranvar") =3D=3D 0)=20 {
           &n= bsp;           =20 ranvar_ =3D (RandomVariable*)=20 TclObject::lookup(argv[2]);
       =             &= nbsp;   =20 return=20 (TCL_OK);
          =      =20 } else if (strcmp(argv[1], "bandwidth") =3D=3D 0)=20 {
           &n= bsp;           =20 bandwidth_ =3D=20 atof(argv[2]);
         &= nbsp;           &n= bsp; =20 return=20 (TCL_OK);
------------------------------------------------------------= -            =    =20
 
}
class LinkDelay : = public Connector=20 {
 public:
       =20 LinkDelay();
        void = recv(Packet* p,=20 Handler*);
        void = send(Packet* p,=20 Handler*);
        void = handle(Event*=20 e);
        double delay() { = return=20 delay_; }
        inline double=20 txtime(Packet* p)=20 {
           &n= bsp;   =20 return (8. * hdr_cmn::access(p)->size() /=20 bandwidth_);
       =20 }
        inline double = txtime2(Packet* p)=20 {
           &n= bsp;   =20 double txt =3D=20 ranvar_->value();
        &= nbsp;      =20 return=20 txt;
           = ;    =20 //return (8. * = hdr_cmn::access(p)->size() /=20 bandwidth_);
       =20 }
        double bandwidth() const = {=20 return bandwidth_; }
        void=20 pktintran(int src, int=20 group);
 protected:
       = int=20 command(int argc, const char*const*=20 argv);
        RandomVariable*=20 ranvar_;
         void=20 reset();
        double=20 bandwidth_;      /* bandwidth of underlying = link=20 (bits/sec) */
        double=20 delay_;          /* line = latency=20 */
        Event=20 intr_;
        int=20 dynamic_;           /* = indicates whether or not link is ~ */
 
        PacketQueue*=20 itq_;
        int=20 total_[4];
};
 
void = LinkDelay::recv(Packet* p,=20 Handler* h)
{
        double = txt =3D=20 txtime2(p);
----------------------------------------------------------------= -----------       =20
 
}
 
in a tcl = file,
 
set dm_ [new = DelayLink]
set rv=20 [new RandomVariable/Uniform]
$rv    =20 set     min_ 0.
$rv    =20 set     max_ 1.
$dm_ ranvar $rv
 
 
 
 
 
 
------=_NextPart_000_0026_01BF6A7E.3115A970-- From L.Wood@eim.surrey.ac.uk Sun Jan 30 12:19:01 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Sun, 30 Jan 2000 12:19:01 +0000 (GMT) Subject: Bad Nam trace format In-Reply-To: <000b01bf6a9a$2cf932a0$16d2523e@fofinha> Message-ID: On Sat, 29 Jan 2000, Claudia Jacy Barenco wrote: > When running Nam I received the following error: > "Nam:badly formatted event at offset 1061 in out.nam" > The line in out.nam is: > "+ 0.288906 0 1 pareto 210 ------- 0 4.0 5.0 0 0" > What is wrong with this format ? at a guess, the name pareto, which nam doesn't recognise. (couldn't find 'pareto' anywhere in a recent build of nam except random.h and rng.h, which are pretty much copies of the same files in ns, and unused by nam itself.) Should an application generating a pareto-like distribution be connected to a traffic agent (UDP/TCP) which nam will recognise? Better approach than the pareto-agent you seem to have there. (You could always lie in your ns code and call it cbr, I suppose.) L. PGP From L.Wood@eim.surrey.ac.uk Sun Jan 30 16:53:22 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Sun, 30 Jan 2000 16:53:22 +0000 (GMT) Subject: Bad Nam trace format In-Reply-To: Message-ID: On Sun, 30 Jan 2000, Lloyd Wood wrote: > On Sat, 29 Jan 2000, Claudia Jacy Barenco wrote: > > > When running Nam I received the following error: > > "Nam:badly formatted event at offset 1061 in out.nam" > > The line in out.nam is: > > "+ 0.288906 0 1 pareto 210 ------- 0 4.0 5.0 0 0" > > What is wrong with this format ? > > at a guess, the name pareto, which nam doesn't recognise. ...since that's obviously .tr format, and this shouldn't be cropping up in a .nam tracefile anyway. Looks like you've got your trace-all and namtrace-all filenames mixed up at some point; the odd thing is that nam would normally complain about these things far earlier than line 1061... bad guess - arghle. (thanks, Miroslav.) L. > (couldn't > find 'pareto' anywhere in a recent build of nam except random.h and > rng.h, which are pretty much copies of the same files in ns, and > unused by nam itself.) > > Should an application generating a pareto-like distribution be > connected to a traffic agent (UDP/TCP) which nam will recognise? > Better approach than the pareto-agent you seem to have there. (You > could always lie in your ns code and call it cbr, I suppose.) > > L. PGP From yanyu@catarina.usc.edu Sun Jan 30 21:53:40 2000 From: yanyu@catarina.usc.edu (Yan Yu) Date: Sun, 30 Jan 2000 13:53:40 -0800 (PST) Subject: add a command in ns In-Reply-To: Message-ID: Hello, all, In ns, when I add a new command(which I can use in Tcl script) for a class, like Node, I add something in Node::command, like: else if (strcmp(argv[1], "setmode") == 0) { mode = atoi( argv[2] ); return (TCL_OK); } well, mode is a global variable I added. It can invoke that piece of code,( I use gdb to see that). but after that, it complains invalid command name: I copy the err msg in gdb as follows, ( when run it outside the debugger, it reports the same error.) ------------------------------- 0x2a423c in Tcl_Main () (gdb) n Single stepping until exit from function Tcl_Main, which has no line number information. invalid command name "" while executing "[$node_(0) setmode 1]" ----------------------------- so I must miss something, I am wondering what else I need to modify to add a new command that I can use in Tcl script. Thanks a lot in advance! yan From intanago@isi.edu Sun Jan 30 22:07:19 2000 From: intanago@isi.edu (Chalermek Intanagonwiwat) Date: Sun, 30 Jan 2000 14:07:19 -0800 (PST) Subject: add a command in ns In-Reply-To: Message-ID: Maybe something overlooked. For example, make sure argc=3. Chalermek On Sun, 30 Jan 2000, Yan Yu wrote: > Hello, all, > > In ns, when I add a new command(which I can use in Tcl script) for a class, > like Node, I add something in Node::command, like: > > else if (strcmp(argv[1], "setmode") == 0) { > > mode = atoi( argv[2] ); > return (TCL_OK); > > } > > well, mode is a global variable I added. > It can invoke that piece of code,( I use gdb to see that). > but after that, it complains invalid command name: > > I copy the err msg in gdb as follows, ( when run it outside the debugger, it > reports the same error.) > ------------------------------- > 0x2a423c in Tcl_Main () > (gdb) n > Single stepping until exit from function Tcl_Main, > which has no line number information. > invalid command name "" > while executing > "[$node_(0) setmode 1]" > > ----------------------------- > so I must miss something, I am wondering what else I need to modify to > add a new command that I can use in Tcl script. > > Thanks a lot in advance! > yan > > From jhbae@cslab5.korea.ac.kr Mon Jan 31 02:24:48 2000 From: jhbae@cslab5.korea.ac.kr (Bae Jin Han) Date: Mon, 31 Jan 2000 11:24:48 +0900 Subject: Questions Message-ID: <3894F270.CF6C889B@cslab5.korea.ac.kr> --------------689EB9FD34C7D30DF2C93927 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I want ns-2.1b6 with differentiated Services Package. On web, I download Release of ns-2.1b6. At ns-2_1b6_tar.tar file, After tar -xvf ns-2_1b6_tar.tar , Error checksum occurred. I don't know cause for it. Why ? Give me Answer. I want to simulation with ns-2.1b6. Regards, Jin Han. --------------689EB9FD34C7D30DF2C93927 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi,
I want ns-2.1b6 with differentiated Services Package.
On web, I download Release of ns-2.1b6.
At ns-2_1b6_tar.tar file,
After tar -xvf ns-2_1b6_tar.tar , Error checksum occurred.
I don't know cause for it.
Why ? Give me Answer.

I want to simulation  with ns-2.1b6.

Regards,
Jin Han. --------------689EB9FD34C7D30DF2C93927-- From tccheng@csis.hku.hk Mon Jan 31 03:39:59 2000 From: tccheng@csis.hku.hk (Cheng Tak Cheung (CE)) Date: Mon, 31 Jan 2000 11:39:59 +0800 (HKT) Subject: Questions In-Reply-To: <3894F270.CF6C889B@cslab5.korea.ac.kr> Message-ID: This file is gzipped. try tar xfvz ns-2_1b6_tar.tar On Mon, 31 Jan 2000, Bae Jin Han wrote: > Hi, > I want ns-2.1b6 with differentiated Services Package. > On web, I download Release of ns-2.1b6. > At ns-2_1b6_tar.tar file, > After tar -xvf ns-2_1b6_tar.tar , Error checksum occurred. > I don't know cause for it. > Why ? Give me Answer. > > I want to simulation with ns-2.1b6. > > Regards, > Jin Han. > From jhbae@cslab5.korea.ac.kr Mon Jan 31 05:45:46 2000 From: jhbae@cslab5.korea.ac.kr (Bae Jin Han) Date: Mon, 31 Jan 2000 14:45:46 +0900 Subject: Question Message-ID: <38952189.C1A48BD2@cslab5.korea.ac.kr> --------------A495F4B3D0BFE6F86232B69B Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Hi, I want to build ns simulator on Windows. But At http://www-mash.cs.berkeley.edu/ns/ns-build.html, For Windows 95/98/NT: (Microsoft Visual C++ is required) A self-extracted ns binary code for windows (32-bit) is available for downloading. I can't download. As example, The requested URL /dist/vint/ns2.1b4.exe was not found on this server. How I get it ? Regards, Jin Han. --------------A495F4B3D0BFE6F86232B69B Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit Hi,

I want to build ns simulator on Windows.
But
At  http://www-mash.cs.berkeley.edu/ns/ns-build.html,
For Windows 95/98/NT: (Microsoft Visual C++ is required)
A self-extracted ns binary code for windows (32-bit) is available for downloading.
I can't download.
As example, The requested URL /dist/vint/ns2.1b4.exe was not found on this server.
How I get it ?

Regards,
Jin Han. --------------A495F4B3D0BFE6F86232B69B-- From swtan@mmu.edu.my Mon Jan 31 16:14:23 2000 From: swtan@mmu.edu.my (Tan Su Wei) Date: Mon, 31 Jan 2000 16:14:23 +0000 Subject: question about debugging ns Message-ID: <3895B4DF.D36D7685@mmu.edu.my> Hi all, I'd read the ns documentation on debugging ns...(I'd never use gdb before) and tried to configure it with --enable-debug and --with-tcldebug , but when i start a script with a line of : debug 1, there is an message saying debug mode was on configure... what i'd done is : in ns-2.1b5 directory: ./configure --enable-debug --with-tcldebug= is there something wrong in what i had done...? and where is the apppropriate place to put the "debug 1" in the script? about c++ level debugging in ns-manual, there is a few line macro for gdb to see what happen in tcl subroutines: define pargvc set $i=0 while $i < argc p argv[$i] ... .. end I'm not sure how to implement this few line in gdb. Thank you. Regards Tan Su Wei From L.Wood@eim.surrey.ac.uk Mon Jan 31 14:33:24 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 31 Jan 2000 14:33:24 +0000 (GMT) Subject: Question In-Reply-To: <38952189.C1A48BD2@cslab5.korea.ac.kr> Message-ID: On Mon, 31 Jan 2000, Bae Jin Han wrote: > I want to build ns simulator on Windows. > But > At http://www-mash.cs.berkeley.edu/ns/ns-build.html, > For Windows 95/98/NT: (Microsoft Visual C++ is required) > A self-extracted ns binary code for windows (32-bit) is available for > downloading. > I can't download. > As example, The requested URL /dist/vint/ns2.1b4.exe was not found on > this server. > How I get it ? This is one reason why the inaccurate berkeley webserver directory should be retired in favour of a proper mirror of the ftp site. Files are missing/in the wrong place/larger than they need to be etc, as noted previously on this list. Look on ftp via the links at the top of the page. L. don't the binary executables need updating for 2.1b6? PGP From 135740@bud.cc.swin.edu.au Sat Jan 29 15:33:41 2000 From: 135740@bud.cc.swin.edu.au (Iyad Rahwan) Date: Sun, 30 Jan 2000 02:33:41 +1100 Subject: compiling an agent example Message-ID: <00c601bf6a6e$39277fe0$6876ba88@186.118.124.swin.edu.au> This is a multi-part message in MIME format. ------=_NextPart_000_00C3_01BF6ACA.6C0A0FC0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable I am trying to try the ping agent example from the ns Tutorial by Marc = Greis. It mentioned that I should do a "make depend" before I do a = "make". Could you please explain to me how I can do a "make depend".. Thanks ------=_NextPart_000_00C3_01BF6ACA.6C0A0FC0 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable

I am trying to try the ping agent example from the = ns Tutorial=20 by Marc Greis. It mentioned that I should do a "make depend" before I do = a=20 "make". Could you please explain to me how I can do a "make=20 depend"..
 
Thanks
------=_NextPart_000_00C3_01BF6ACA.6C0A0FC0-- From nemo@comnets.uni-bremen.de Mon Jan 31 16:22:49 2000 From: nemo@comnets.uni-bremen.de (Niko A. Fikouras) Date: Mon, 31 Jan 2000 17:22:49 +0100 Subject: Calling C++ functions at the Tcl front end Message-ID: <3895B6D9.6FBB97A0@comnets.uni-bremen.de> Hi, my question is the following. I have a model that is consisted of several C++ functions. All of these functions are called by the model when a certain event occurs i.e. the receipt of a packet. But, I wish to place a call to a certain function explicitly at certain time points. How could this be made. Thanks in advance, --Niko From Tarik Alj Mon Jan 31 17:00:01 2000 From: Tarik Alj (Tarik Alj) Date: Mon, 31 Jan 2000 12:00:01 -0500 (EST) Subject: Calling C++ functions at the Tcl front end Message-ID: <200001311700.MAA25636@someware.INRS-Telecom.UQuebec.CA> double time = Scheduler::instance().clock(); at the C++ level $ns at $time [$model call-function] from the command function of model at the Tcl level the command method is used to make C++ calls from OTcl; like in: $mac target $ll for example. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Mon, 31 Jan 2000 17:22:49 +0100 > From: "Niko A. Fikouras" > X-Accept-Language: en > MIME-Version: 1.0 > To: ns-users@mash.cs.berkeley.edu > Subject: Calling C++ functions at the Tcl front end > Content-Transfer-Encoding: 7bit > > Hi, > > my question is the following. I have a model that is consisted of > several C++ functions. > All of these functions are called by the model when a certain event > occurs i.e. the receipt > of a packet. But, I wish to place a call to a certain function > explicitly at certain time points. > > How could this be made. > > Thanks in advance, > > --Niko Tarik From mark_wibble@hotmail.com Mon Jan 31 17:01:55 2000 From: mark_wibble@hotmail.com (Mark Butler) Date: Mon, 31 Jan 2000 09:01:55 PST Subject: compiling an agent example Message-ID: <20000131170155.97530.qmail@hotmail.com> Just type in 'make depend' in the ns directory - it saves having to 'make clean' then re-make the whole of ns for certain changes >From: "Iyad Rahwan" <135740@bud.cc.swin.edu.au> >To: >Subject: compiling an agent example >Date: Sun, 30 Jan 2000 02:33:41 +1100 > >I am trying to try the ping agent example from the ns Tutorial by Marc >Greis. It mentioned that I should do a "make depend" before I do a "make". >Could you please explain to me how I can do a "make depend".. > >Thanks ______________________________________________________ Get Your Private, Free Email at http://www.hotmail.com From czermin@dune.LKN.E-Technik.TU-Muenchen.DE Mon Jan 31 17:22:33 2000 From: czermin@dune.LKN.E-Technik.TU-Muenchen.DE (Czermin Martin) Date: Mon, 31 Jan 2000 18:22:33 +0100 Subject: Question about WFQ Message-ID: <3895C4D9.70998D98@lkn.ei.tum.de> Hi all, I have had a Q about the differences of WFQ and other scheduling algorithm like DRR or WRR some days ago. To simulate the Delays in an integrated service environment I need to implement one of the algorithm mentioned above. As far as I informed it磗 the best way to use an WFQ scheduler. Does anybody know if there is already an implementation in the new version 2.1b6 of ns. Can I also simulate with other schedulers like DRR or WRR and are the results compareable to the results I would get if I use WFQ? Thanks in advance Martin From agandhi@hopper.unh.edu Mon Jan 31 18:37:10 2000 From: agandhi@hopper.unh.edu (Arun Gandhi) Date: Mon, 31 Jan 2000 13:37:10 -0500 (EST) Subject: Question about finding end to end delay Message-ID: Dear ns-users, I am trying to find the end to end delay of a packet in the network.I am writing a function which takes the trace file as an argument and does the job. I am not sure whether there are some already existing parameters in the core and if i could just use them.In my search and understanding i couldn't find any. Your suggestions are called for. Thanks, Arun ---------------------------------------------------------------------------- Arun Gandhi M.S. Computer Science Office: 603 - 862 - 3786 University of New Hampshire Home : 603 - 862 - 7614 Durham,NH 03824 Email : agandhi@cisunix.unh.edu From L.Wood@eim.surrey.ac.uk Mon Jan 31 18:43:42 2000 From: L.Wood@eim.surrey.ac.uk (Lloyd Wood) Date: Mon, 31 Jan 2000 18:43:42 +0000 (GMT) Subject: Question about finding end to end delay In-Reply-To: Message-ID: On Mon, 31 Jan 2000, Arun Gandhi wrote: > I am trying to find the end to end delay of a packet in the network.I am > writing a function which takes the trace file as an argument and does the > job. > I am not sure whether there are some already existing parameters in the > core and if i could just use them.In my search and understanding i > couldn't find any. > Your suggestions are called for. Read: http://www-mash.cs.berkeley.edu/dist/archive/ns-users/9907/0350.html which does it for you. L. PGP From guille@iies.es Mon Jan 31 17:04:01 2000 From: guille@iies.es (Guillermo Rodriguez Garcia) Date: Mon, 31 Jan 2000 18:04:01 +0100 Subject: compiling an agent example In-Reply-To: <00c601bf6a6e$39277fe0$6876ba88@186.118.124.swin.edu.au> Message-ID: <3.0.3.32.20000131180401.00843510@mailhost.iies.es> At 02:33 30/01/00 +1100, you wrote: > I am trying to try the ping agent example from the ns Tutorial by Marc >Greis. It mentioned that I should do a "make depend" before I do a "make". >Could you please explain to me how I can do a "make depend".. Thanks You go into the ns source directory and then type (surprise) make depend and hit return. Regards, G. -- Guillermo Rodriguez Garcia From Tarik Alj Mon Jan 31 19:04:22 2000 From: Tarik Alj (Tarik Alj) Date: Mon, 31 Jan 2000 14:04:22 -0500 (EST) Subject: Question about finding end to end delay Message-ID: <200001311904.OAA04534@someware.INRS-Telecom.UQuebec.CA> Here is a way of doing it with tr files (obtained with trace-all) : use awk to filter all the "r" traces at src node and dst node; then using the uid of each packet to discriminate between them you can compute e2e delay by doing a simple substraction. > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to owner-ns-users@irtf.CS.Berkeley.EDU using -f > Date: Mon, 31 Jan 2000 13:37:10 -0500 (EST) > From: Arun Gandhi > To: ns-users@mash.cs.berkeley.edu > Subject: Question about finding end to end delay > MIME-Version: 1.0 > > Dear ns-users, > I am trying to find the end to end delay of a packet in the network.I am > writing a function which takes the trace file as an argument and does the > job. > I am not sure whether there are some already existing parameters in the > core and if i could just use them.In my search and understanding i > couldn't find any. > Your suggestions are called for. > Thanks, > Arun > > ---------------------------------------------------------------------------- > Arun Gandhi > M.S. Computer Science Office: 603 - 862 - 3786 > University of New Hampshire Home : 603 - 862 - 7614 > Durham,NH 03824 Email : agandhi@cisunix.unh.edu Tarik Alj INRS-Telecommunications Place Bonaventure 900 De La Gauchetierre Ouest Niveau C, Case Postale 644 Montreal, Qc, H5A 1C6 Canada From bagula@cs.sun.ac.za Mon Jan 31 19:31:34 2000 From: bagula@cs.sun.ac.za (Antoine Bagula) Date: Mon, 31 Jan 2000 21:31:34 +0200 (SAST) Subject: Help Message-ID: Dear all I am a new user of NS which I should use in my research work. I have found Marc Greis tutorial very interesting as a starting point for my NS work. I have just been able to install NS and run it with NAM. But I still have some problems on how to obtain animated results from NS or any other results. I am still just able to run examples in Greis'tutorial but not to receive any results from the pictures drawn on a NAM screen. I may be missing something. As a case , I am not able to see any result from his example (see tutorial V.4) concerning "Monitoring a Queue" : any colors or packets in the queue. Can someone help by solving that problem and/or give some other links for a beginner like me ? Regards Bagula. ############################################################################### BIGOMOKERO ANTOINE BAGULA Home : Gardenia 2 Dept of Computer Science Skyvuedrive University of Stellenbosch Kluis Rivier Private Bag XI 7580 7602 Matieland Cape Town South Africa South Africa Email :bagula@cs.sun.ac.za bagula@unikin.net http://www.cs.sun.ac.za/projects/COE/#members http://www.cs-net.org Fax : +2721 8084416 (W) Tel : +2721 8084605 (W) Cell :+2782 9760453 ############################################################################### From aroetter@CS.Stanford.EDU Mon Jan 31 22:05:15 2000 From: aroetter@CS.Stanford.EDU (Alex Roetter) Date: Mon, 31 Jan 2000 14:05:15 -0800 (PST) Subject: flow id's and .tr files Message-ID: i'm having trouble setting flow id's for cbr senders attached to udp agents, and seeing that flow id appear in the tracefile. if i set the fid_ variable of a tcp agent, it works no problem (the new flow id is shown in the trace file for packets from that agent) but if i do the same thing for a udp agent $udp set fid_ 7 the flow id still shows up as zero in the trace file. this makes it very differnet to differentiate packets by sender for post-processing of the .tr files. does anyone know how to do this? thanks, alex From agandhi@cisunix.unh.edu Mon Jan 31 22:50:50 2000 From: agandhi@cisunix.unh.edu (Arun Gandhi) Date: Mon, 31 Jan 2000 17:50:50 -0500 Subject: Question about finding end to end delay References: <200001311904.OAA04534@someware.INRS-Telecom.UQuebec.CA> Message-ID: <389611CA.51DBCE73@cisunix.unh.edu> --------------0BE1C7DEBEF1FE7AF3EFFD5F Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit This is what i am thinking. Thanks. Arun Tarik Alj wrote: > Here is a way of doing it with tr files (obtained with trace-all) : use awk to > filter all the "r" traces at src node and dst node; then using the uid of each > packet to discriminate between them you can compute e2e delay by doing a simple > substraction. > > > X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to > owner-ns-users@irtf.CS.Berkeley.EDU using -f > > Date: Mon, 31 Jan 2000 13:37:10 -0500 (EST) > > From: Arun Gandhi > > To: ns-users@mash.cs.berkeley.edu > > Subject: Question about finding end to end delay > > MIME-Version: 1.0 > > > > Dear ns-users, > > I am trying to find the end to end delay of a packet in the network.I am > > writing a function which takes the trace file as an argument and does the > > job. > > I am not sure whether there are some already existing parameters in the > > core and if i could just use them.In my search and understanding i > > couldn't find any. > > Your suggestions are called for. > > Thanks, > > Arun > > > > ---------------------------------------------------------------------------- > > Arun Gandhi > > M.S. Computer Science Office: 603 - 862 - 3786 > > University of New Hampshire Home : 603 - 862 - 7614 > > Durham,NH 03824 Email : agandhi@cisunix.unh.edu > > Tarik Alj > > INRS-Telecommunications > Place Bonaventure > 900 De La Gauchetierre Ouest > Niveau C, Case Postale 644 > Montreal, Qc, H5A 1C6 > Canada -- Arun Gandhi M.S. (Computer Science) Phone : 603-862-7614 (Res) University of New Hampshire, 603-862-3786 (Off) Durham, NH 03824 Email : agandhi@cisunix.unh.edu --------------0BE1C7DEBEF1FE7AF3EFFD5F Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit This is what i am thinking. Thanks.
Arun

Tarik Alj wrote:

Here is a way of doing it with tr files (obtained with trace-all) : use awk to
filter all the "r" traces at src node and dst node; then using the uid of each
packet to discriminate between them you can compute e2e delay by doing a simple
substraction.

> X-Authentication-Warning: mash.CS.Berkeley.EDU: majrdomo set sender to
owner-ns-users@irtf.CS.Berkeley.EDU using -f
> Date: Mon, 31 Jan 2000 13:37:10 -0500 (EST)
> From: Arun Gandhi <agandhi@hopper.unh.edu>
> To: ns-users@mash.cs.berkeley.edu
> Subject: Question about finding end to end delay
> MIME-Version: 1.0
>
> Dear ns-users,
> I am trying to find the end to end delay of a packet in the network.I am
> writing a function which takes the trace file as an argument and does the
> job.
> I am not sure whether there are some already existing parameters in the
> core and if i could just use them.In my search and understanding i
> couldn't find any.
> Your suggestions are called for.
> Thanks,
> Arun
>
> ----------------------------------------------------------------------------
> Arun Gandhi
> M.S. Computer Science                        Office: 603 - 862 - 3786
> University of New Hampshire                  Home  : 603 - 862 - 7614
> Durham,NH 03824                              Email : agandhi@cisunix.unh.edu

Tarik Alj

INRS-Telecommunications
Place Bonaventure
900 De La Gauchetierre Ouest
Niveau C, Case Postale 644
Montreal, Qc, H5A 1C6
Canada

-- 
Arun Gandhi                                                
M.S. (Computer Science)                                 Phone : 603-862-7614 (Res)
University of New Hampshire,                                    603-862-3786 (Off)
Durham, NH 03824                                        Email : agandhi@cisunix.unh.edu
  --------------0BE1C7DEBEF1FE7AF3EFFD5F-- From Martin Koyabe Mon Jan 31 23:32:15 2000 From: Martin Koyabe (Martin Koyabe) Date: Mon, 31 Jan 2000 23:32:15 +0000 (GMT) Subject: Debugger !! Message-ID: <200001312332.XAA06179@churchward.erg.abdn.ac.uk> L, Just a small one here !! I have installed the debugger correctly (ns using --with-tcldebug) with add "debug 1" line in my TCL code running OK. But when I run the tcl-code I get this at the "debug 1" line --- ns mftp_sim_20.tcl Creating links for WAN .... Creating links for LAN with (1..10) nodes... Creating links for LAN with (10..20) nodes... Inserting link errors (1..10) nodes... Inserting link errors (10..20) nodes... 2: lappend auto_path $dbg_library dbg2.0> --- Is there some library location I need to specify !! The ns-doc doesn't really explain this bit !! Tx -- Martin From huangpeng_buaa@a-1.net.cn Tue Jan 11 02:01:28 2000 From: huangpeng_buaa@a-1.net.cn (HUANG PENG) Date: Tue, 11 Jan 2000 10:01:28 +0800 Subject: How many Biconnected components? Message-ID: <003301bf5bd7$c6d4e300$7b8370ca@buaa.edu.cn> This is a multi-part message in MIME format. ------=_NextPart_000_002D_01BF5C1A.D4410800 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 SGksZXZlcnlvbmU6DQogICAgDQogICAgSSBoYXZlIHJlYWQgdGhlIGd0LWl0bSBwYXBlcnMuIEFj Y29yZGluZyB0byB0aGF0IHBhcGVyIHRoZXJlIHNob3VsZCBiZSBuIGJpY29ubmVjdGVkIGNvbXBv bmVudHMgaW4gdGhlIHN0YXIgdG9wb2xvZ3kgYW5kIHRoZSBjaGFpbiB0b3BvbG9neS4gSXMgdGhh dCByaWdodD8gSSB0aGluayB0aGUgbnVtYmVyIG9mIHRoZSBiaWNvbm5lY3RlZCBjb21wb25lbnRz IHNob3VsZCBiZSAobi0xKS4gQXR0YWNoZWQgaW4gdGhpcyBtYWlsIHRoZXJlIGFyZSBhIHBpY3R1 cmUgb2YgdGhlIHN0YXIgYW5kIGNoYWluIHRvcG9sb2d5LiBJIGNvdW50IEEtQiwgQi1DLCBDLUQs IEQtRSwgRS1GIGFyZSB0aGUgYmljb25uZWN0ZWQgY29tcG9uZW50cyBpbiB0aGUgY2hhaW4gdG9w b2xvZ3kgYW5kIHRoZSBBLUksIEItSSwgQy1JLCBELUksIEUtSSwgRi1JLCBHLUkgYXJlIHRoZSBi aWNvbm5lY3RlZCBjb21wb25lbnRzLCBib3RoIGFjY291bnRzIGZvciAobi0xKS4gSXMgdGhlcmUg YW55IG1vcmU/IGNhbiB5b3UgdGVsbCBtZSBpZiBJIGFtIHJpZ2h0Pw0KDQogICAgICAgVGhhbmsg eW91IGluIGFkdmFuY2UhDQoNCg0KDQo= ------=_NextPart_000_002D_01BF5C1A.D4410800 Content-Type: application/x-zip-compressed; name="question.zip" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="question.zip" UEsDBBQAAAAIAKWpSyIt7seGrgsAAL5MBAAMAAAAcXVlc3Rpb24uYm1w7dhNjuy4EYVRNeBxj70E Tzx1zz229+VVeeR9qcsvUYJKlEiKEskbEd8B3kNlJqUMRvBm/fzz3//911+W//vH17+/ff3747dl +fvXv9+Wv/56/j9fr//v9+XXv28rAG3kFNBHTgF95BTQR06D4M8RpjG4OJi1XcwuDmZtF7OLg1nb xezi2GbN0M1hZHGQU7sYWRzLTnk1lDCyOPh+ahcji4Oc2sXI4jjMmtEbwrCCWBLlayCDeQH6yCmg j5w6k/58y4gdYIhuZCJJWq1jfD7UzJFZ28XsHKgfIuM2isFlmPj97m55TNwipnZl3xnlLpHTCJja FRM5bSuMoZvDyE4d2kJOMRcjO3X4Zir7W2pzVczdFuZ1ykpbyGkQzOuUg5y2vQRNjOwUOYUURnbl 8CtqZuVcV7XdfR7KmNoVcgodTC1j2Smvnue0vPonoY/BOUBO3WNwPqRzrHkGVjA7Nw6jzD+ELYzP k/2v0vsvmLJ1TLDGcqF85QwWixSvdjo6k5c/POJBKC+awW5LJ6InGZXN0eyh6ao4lgc05Ep9Z8hp PetdnYVunLrbFsE2OiiJw7mhFaca2qLWSd85XRJXK30gp6nmnkg108fnRj6qlSsdIKcpctoDOX2C nKbI6et6tDTU0SWnKXL6uq45jXCGyWkq3xMrn/BWfu593m1yGlOxJ21HbjATOW1u9eelzdUaN8jp qYbzY+X710SHkp6EdP8qOQ2rpi13T91g4jlt6PDVq+Q0slsHSf+bl4hPVc9Dmi7wfZLJ6ZXKziy/ lNcNJ1tVfWPzrx5kFjtATjNqmkNOb6lsF8fygIbkFfuzVP8gN5hsSZUtxR49KVqy9ssyNxlPuZ7K lmJDW14k1UyK8YQGvkunn1TiCT18nUhLKcMT2tiDQlepwRM62cn0xlKAJzSzn7m9jfzu/tDPria2 N+Zbe0VLe5vV4Wjv6xtdHWBKk+O8aQQ0dozxfY7wjnHQ22EGt9r320VDe0ca2W2v7xUTHR5sWMP9 vVFkNHm8MT339C6gz1MMaLuPt8AHrZ6ld+et3x97dHuirs23e3OkaPhc/fpv8c64Qs+nW34pr7up 0z05MFPQdhGvD0L/hqhH83W8Owvlu+Eu+i/lxXFo3gptGIGatyaidh88wRQEvTIUnZvgOQah6flc FO6AtzALWQ9HM/dyvItxKHsynVnXogcmIq55QOMvRD8MRV/bjEZehd6YiwkNYxpzCcZgNFbcnVTv 9RiJ6Rhya1j9FmM8BmRL/bx6rMQszMicypG9uwxzMSaLaqb21hooYFJGFQf3fAF0MCy78rN78irU BJzXslNere1qC/kN+ti4myHWCJXTdKwOBn21o+3/dMFqmcshFsX5OMrs1PqgTyN5+Hpbk15uhe8h 5kX4IKrco91WnO7x9BnTeywv8ptW3zm9uzuL3dhqPhR/9dD0Hiv5O9Xk9MBcQ/YF3/3ahAhDLHKc 0+atGepJWmr6fTOzxoS2ap0dbHKaMp3T7cn9/6cLrCCnKzk9Yz2nn+c/Tl9d7ewxwhBrxMzpAheu 5ruSUzuat2aoJ1elFk+ylT1GGGINcpqyntOF30/JqR0xc7o9k36RrjGBnK6uc7o27c5cQ/YF3/3a hAhDLCKnB3aP8aHyq4fk1CKjO1q+nT48rEyfvGK6G+mTNctMiDDEPLubOlR+tZH6Dbppxd0nTYgw xwz3OV3rvo9Y70O6x/ThtnK1qWaILkO6xsjptiBds3w7vUTfVRj3L6UbNL3fzBxPL/HB0xG9Wrm3 /FS+QNhV/TV7dLB3N3OsYXePS6J8jSPF/UZriG+mc5p56FvNZivXhP2gs4WcmlO50/qGkFB95NSW Hukjp/rIqSG39khOPTE6o8OvVIeHLt3dne+cLj+VLzCOz1ITGsbkO6frz7LdH2Nyqq9tRqFymj50 hpyKax4QOfWEnCp7Mh1y6gk5lfVwNOTUE3Kq6flcau6w/FRcL4WcYq5XhuJ+suQUE701EXLqCTmV 8uI4QuU02ocSJnp3Fsp3e275qXyBceRUxOuD0L8h6tF8BT2mYOWeqEHnp+s0Alu3RR5tn6tf/y3e GVfo+URdm2/35kjR8Fl6d976/bFHt6cY0HYfb4EPWj3emJ57ehfQ58GGNdzfG0VGk0ca2W2v7xUT HR5mcKt9v100tHeM8X2O8I5x0NsBpjQ5zptGQGN7m9XhaO/rG13tamJ7Y761V7S0n7m9jfzu/tDP TqY3lgI8oZk9KHSVGjyhk68TaSlleEIb36XTTyrxhB6+SKqZFOMJDayxXDisubp8CvF6alqKDZ3J yx+e7VXBNsqWVNlS7NGTjMrmaPbQdFUcywMacqX+RJHTevXt4mTu0Y1Td88SP/fWuNsuDueGVpyq acthjVonxT86GjocGa1INR8hqWbqf24Q1Xr0IVXsydUCcprR9snG+fygD6l8TzKvktMrzU3jfH7Q h5SPQ+WjGM7nB31I+ThUJn7uLeJ8ftCHFDntgZw+QR9ONbTFRy66ulsSh3NDK06R0x7IaTNaceVW ZxyEYoz6qjiZe3TjivUTRU49oRsZNc2RbaDpwjiWBzQkr9gfcnqX3ZZORE+Klqzy9TNYKe9U+fp4 aIsz21Hf/mfEDjBENw6RPHzNoE1jfD6kc6x5BlYwOwdOh1j/JPQxOOuuJnj3eSiLMLXl2+lD6xzk 1NlEegjSnMM23Ye0+aXx9sUQ1Svk1DRnOU0f4iNgTj1tuTmMIk0glZXiNGrZKa82gpwGESqnhy8c aNuLTgfIaaVoOXW2X3IaBDm17u6OpDpATiuFapSzX04/TOd0Teohuaei5bS8yKD6fQl2gJzWIKcO mM7p+v1zzqZ8QTy0xYeaOTJru5idG/lRMmjTGJ8ny7XyxRDGBAF95BTQR04BfeQU0EdOAX3k1Bb+ ihsTszZnPzLGFwSDNucwMiYYAVM2h5wGxJTNIacBMWVz+P00IAZtDn/vDYhZm8P304AYtDnkNCAG bQ45DYhB27L8VL4ALjBrQB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0kVNAHzkF9JFT QB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0kVNA HzkF9JFTQB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0Af OQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0kVNAHzkF9JFTQB85BfSRU0AfOQX0VeZ0uVC+cjjrpZYv G+60Tkp96FapxS3ktynVBB+l0tI2vkvNF1+5NYUOOCuVlt7ivtR8rq9eOpi+fUrtgVJ7aCv16qq7 25m4fUrtgVJ7aC71rZyu87bPiXqd4+mvNks9vbZtI1O2T6k9UOrrmt/xcyE5HYZSeyCnDcxtfxha 2kOcUl/P6Xjlss6U7/u2ck1nyvftoFzWmfJ9OyiXdaZ83w7KZZ1Z3s5pcyVPUGoPlPo6clpelKDU PEp9XY+cZp7PmDKm9f77zqpztVOq4+mvNkslpyOR09cFz2n+pdTEw7/eefe5da52SnU5/dVsqa/k lMNfj1J7cF9q/qqae07f+Ael9kCpPTSUWrzk+YJhip85hkrNvDpSsWlWWlqzYJiGUiujnVG8fCQr da6U2kGmTuulqtUPIEVOAX3kFNBHTgF95BTQl8/p4Q9Qmb9HSfn5xzPRak0000QnU0uifM1ADVUV Fx8WWBnWvk7BmsXL2zNU6p74ub1bT4Scpg+nM9RVDkAPXXPKx+krDLVx9ZJTNa/n9LNmU1wsgpy+ xXpONQu+W15lTg9f6FP+MYCcDhA2p+ZO16a8eixyOoDyAVjvl1ezbLGZ09OvFZDTAbY6ZXN6+CKv fhk5fZGhw09Oe+iX0/IiJeT0LeKdvCLe4bvluczp8lP5ghkosp8lUb5moIby1LYAIEVOAX3kFNBH TgF95BTQ95XTPwFQSwECFAAUAAAACAClqUsiLe7Hhq4LAAC+TAQADAAAAAAAAAAAACAAtoEAAAAA cXVlc3Rpb24uYm1wUEsFBgAAAAABAAEAOgAAANgLAAAAAA== ------=_NextPart_000_002D_01BF5C1A.D4410800-- From huangpeng_buaa@a-1.net.cn Tue Jan 11 10:49:22 2000 From: huangpeng_buaa@a-1.net.cn (HUANG PENG) Date: Tue, 11 Jan 2000 18:49:22 +0800 Subject: How many Biconnected components? Message-ID: <002101bf5c21$859d06c0$7b8370ca@buaa.edu.cn> This is a multi-part message in MIME format. ------=_NextPart_000_001F_01BF5C64.930F4640 Content-Type: text/plain; charset="gb2312" Content-Transfer-Encoding: base64 SGksZXZlcnlvbmU6DQogICAgDQogICAgSSBoYXZlIHJlYWQgdGhlIGd0LWl0bSBwYXBlcnMuIEFj Y29yZGluZyB0byB0aGF0IHBhcGVyIHRoZXJlIHNob3VsZCBiZSBuIEJpY29ubmVjdGVkIGNvbXBv bmVudHMgaW4gdGhlIHN0YXIgdG9wb2xvZ3kgYW5kIHRoZSBjaGFpbiB0b3BvbG9neSwgYm90aCBv ZiB3aGljaCBjb250YWlucyBuIG5vZGVzLiBJcyB0aGF0IHJpZ2h0PyBJIHRoaW5rIHRoZSBudW1i ZXIgb2YgdGhlIGJpY29ubmVjdGVkIGNvbXBvbmVudHMgc2hvdWxkIGJlIChuLTEpLiBBdHRhY2hl ZCBpbiB0aGlzIG1haWwgdGhlcmUgaXMgYSBwaWN0dXJlKHppcHBlZCkgb2YgdGhlIHN0YXIgYW5k IGNoYWluIHRvcG9sb2d5LiBJIGNvdW50IEEtQiwgQi1DLCBDLUQsIEQtRSwgRS1GIGFyZSB0aGUg Ymljb25uZWN0ZWQgY29tcG9uZW50cyBpbiB0aGUgY2hhaW4gdG9wb2xvZ3kgYW5kIHRoZSBBLUks IEItSSwgQy1JLCBELUksIEUtSSwgRi1JLCBHLUkgYXJlIHRoZSBiaWNvbm5lY3RlZCBjb21wb25l bnRzIGZvciB0aGUgc3RhciB0b3BvbG9neSwgYm90aCBhY2NvdW50cyAobi0xKS4gSXMgdGhlcmUg YW55IG1vcmU/IFdobyBjYW4gdGVsbCBtZSBpZiBJIGFtIHJpZ2h0Pw0KDQogICAgICAgVGhhbmsg eW91IGluIGFkdmFuY2UhDQo= ------=_NextPart_000_001F_01BF5C64.930F4640 Content-Type: application/x-zip-compressed; name="question.zip" Content-Transfer-Encoding: base64 Content-Disposition: attachment; filename="question.zip" UEsDBBQAAAAIAKWpSyIt7seGrgsAAL5MBAAMAAAAcXVlc3Rpb24uYm1w7dhNjuy4EYVRNeBxj70E Tzx1zz229+VVeeR9qcsvUYJKlEiKEskbEd8B3kNlJqUMRvBm/fzz3//911+W//vH17+/ff3747dl +fvXv9+Wv/56/j9fr//v9+XXv28rAG3kFNBHTgF95BTQR06D4M8RpjG4OJi1XcwuDmZtF7OLg1nb xezi2GbN0M1hZHGQU7sYWRzLTnk1lDCyOPh+ahcji4Oc2sXI4jjMmtEbwrCCWBLlayCDeQH6yCmg j5w6k/5