From srilekhk@nunki.usc.edu  Sat Apr  5 08:22:25 2003
From: srilekhk@nunki.usc.edu (srilekhk)
Date: Sat, 05 Apr 2003 00:22:25 -0800 (PST)
Subject: [Csci551-talk] Regarding Homework 2
Message-ID: <Pine.GSO.4.33.0304050019010.29944-100000@nunki.usc.edu>

Hi,
   For the part of the second question that needs ns-nam, will the TA be
providing the path to ns on our unix accounts? Or should we install ns on
our own accounts? I believe we have a space limitation of 100 MB on our
solaris accounts and the distribution says it needs 250MB of disk space
for the entire ns install. Please do let me know

Thanks and Regards
Srilekha


From shshanks@aludra.usc.edu  Sat Apr  5 14:19:51 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Sat, 05 Apr 2003 06:19:51 -0800 (PST)
Subject: [Csci551-talk] Regarding Homework 2
In-Reply-To: <Pine.GSO.4.33.0304050019010.29944-100000@nunki.usc.edu>
Message-ID: <Pine.GSO.4.33.0304050615560.13127-100000@aludra.usc.edu>

On Sat, 5 Apr 2003, srilekhk wrote:

> Hi,
>    For the part of the second question that needs ns-nam, will the TA be
> providing the path to ns on our unix accounts? Or should we install ns on
> our own accounts?

ns and nam are already installed in the CS551 course account.
Can directly type in '~csci551/ns filename.tcl' and '~csci551/nam
filename.nam' at the aludra/nunki prompt.

> I believe we have a space limitation of 100 MB on our
> solaris accounts and the distribution says it needs 250MB of disk space
> for the entire ns install. Please do let me know

The space limit is 50 MB.

-Sharma


From srilekhk@usc.edu  Sat Apr  5 16:42:41 2003
From: srilekhk@usc.edu (srilekha krishnamurthy)
Date: Sat, 05 Apr 2003 08:42:41 -0800
Subject: [Csci551-talk] Regarding Homework 2
Message-ID: <17c5a3f17c73db.17c73db17c5a3f@usc.edu>

Srilekha Krishnamurthy
Graduate Student 
Computer Science Department
USC

----- Original Message -----
From: Shshank Sharma <shshanks@usc.edu>
Date: Saturday, April 5, 2003 6:19 am
Subject: Re: [Csci551-talk] Regarding Homework 2

> 
> On Sat, 5 Apr 2003, srilekhk wrote:
> 
> > Hi,
> >    For the part of the second question that needs ns-nam, will the TA be
> > providing the path to ns on our unix accounts? Or should we install ns on
> > our own accounts?
> 
> ns and nam are already installed in the CS551 course account.
> Can directly type in '~csci551/ns filename.tcl' and '~csci551/nam
> filename.nam' at the aludra/nunki prompt.
---Thanks for this information. Will try it.

> 
> > I believe we have a space limitation of 100 MB on our
> > solaris accounts and the distribution says it needs 250MB of disk space
> > for the entire ns install. Please do let me know
> 
> The space limit is 50 MB.
----Well, I take the 571 class and students in that class usually have 100 MB space :-) But you are right usually it is 50MB.
> 
> -Sharma
> 


 
 



From xiw@usc.edu  Sun Apr  6 05:43:03 2003
From: xiw@usc.edu (Xi Wang)
Date: Sat, 05 Apr 2003 20:43:03 -0800
Subject: [Csci551-talk] Regarding Homework 2
References: <17c5a3f17c73db.17c73db17c5a3f@usc.edu>
Message-ID: <000c01c2fbf7$03899f80$bbe07d80@xwl3>

I got a segmentation fault for hw2q2b.tcl. The tcl file was not modified
and this happended on both of the aludra and the PC version of ns. What
could be the problem?

Thanks,

Xi


From gnawali@usc.edu  Mon Apr  7 08:06:41 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Mon, 07 Apr 2003 00:06:41 -0700
Subject: [Csci551-talk] compiling proja sample code
Message-ID: <200304070706.h3776gD1022468@enl.usc.edu>

Some students were having difficulty compiling the sample code for
proja on Solaris. I grabbed the source from the course webpage and
made the following changes to get it to compile on aludra (Solaris):

* explicitly add includes <algorithm> and <function.h>
* get rid of ifndef OSTYPE in Makefile

You can download the code that compiles on both platforms from the TA
webpage.

Some of you were also wondering if you will get points taken off if
you decide to not use our code. We are providing you sample code as a
reference implementation of proja. You are not required to use it and
you will not be penalized for not using it.

From gnawali@usc.edu  Mon Apr  7 08:13:22 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Mon, 07 Apr 2003 00:13:22 -0700
Subject: [Csci551-talk] Regarding Homework 2
In-Reply-To: Your message of Sat, 05 Apr 2003 20:43:03 -0800
Message-ID: <200304070713.h377DMf3022660@enl.usc.edu>

I just ran both the files (as it is on the webpage) on aludra and it
worked fine.

you wrote:
> I got a segmentation fault for hw2q2b.tcl. The tcl file was not modified
> and this happended on both of the aludra and the PC version of ns. What
> could be the problem?
> 
> Thanks,
> 
> Xi

From gnawali@usc.edu  Mon Apr  7 09:22:27 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Mon, 07 Apr 2003 01:22:27 -0700
Subject: [Csci551-talk] hw2 more info
Message-ID: <200304070822.h378MRi3023434@enl.usc.edu>

Dear class,

Your answer for HW2 2.b and 2.h should include an excerpt from the
tracefile that you generate. Things to keep in mind:

* It is not enough to just slap a portion of your tracefile in the
answer and expect to get a full credit. Please explain what is in that
excerpt.

* Please do not include your entire tracefile. It will be too long for
me to read it all. Please include as few lines as is needed to support
your answer. I would be surprised if you had to include more than 5
lines of trace. The trace excerpt that you include in your answer does
not have to be contiguous -- feel free to delete the lines that are
irrelevant to your answer.

- om_p

From akulkarn@usc.edu  Mon Apr  7 17:02:20 2003
From: akulkarn@usc.edu (anupam kulkarni)
Date: Mon, 07 Apr 2003 09:02:20 -0700
Subject: [Csci551-talk] Re: Csci551-talk digest, Vol 1 #122 - 1 msg
Message-ID: <1c6a7431c65742.1c657421c6a743@usc.edu>


if anyone is getting a seg fault while running nam it is probably because you are running it over telnet.
try using Xwin. 
Anupam

> I got a segmentation fault for hw2q2b.tcl. The tcl file was not modified
> and this happended on both of the aludra and the PC version of ns. What
> could be the problem?
> 
> Thanks,
> 
> Xi
> 
> 
> 
> --__--__--
> 
> _______________________________________________
> Csci551-talk mailing list
> Csci551-talk@mailman.isi.edu
> http://mailman.isi.edu/mailman/listinfo/csci551-talk
> 
> 
> End of Csci551-talk Digest
> 


From johnh@ISI.EDU  Mon Apr  7 18:46:56 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 07 Apr 2003 10:46:56 -0700
Subject: [Csci551-talk] compiling proja sample code
In-Reply-To: <200304070706.h3776gD1022468@enl.usc.edu>
Message-ID: <200304071746.h37HkuOA009043@dash.isi.edu>

On Mon, 07 Apr 2003 00:06:41 PDT, Omprakash Gnawali wrote: 
>
>Some students were having difficulty compiling the sample code for
>proja on Solaris. I grabbed the source from the course webpage and
>made the following changes to get it to compile on aludra (Solaris):
>
>* explicitly add includes <algorithm> and <function.h>
>* get rid of ifndef OSTYPE in Makefile
>
>You can download the code that compiles on both platforms from the TA
>webpage.

Thanks for fixing those portability buglets, Om.

>Some of you were also wondering if you will get points taken off if
>you decide to not use our code. We are providing you sample code as a
>reference implementation of proja. You are not required to use it and
>you will not be penalized for not using it.

I would like to add to this: your choice of Project A implementation
will not affect your Project B grade in any way.  (Assuming you do a
good job on the Project B parts :-).

   -John Heidemann

From shetye@usc.edu  Mon Apr  7 22:54:41 2003
From: shetye@usc.edu (siddharth shetye)
Date: Mon, 07 Apr 2003 14:54:41 -0700
Subject: [Csci551-talk] project A solution executing ?
Message-ID: <1eb2c361eb293e.1eb293e1eb2c36@usc.edu>

Hi,
I just downloaded the solution to project A from the TA's website and compiled it. But when I test it with painput1 thru painput4 (also recently downloaded from the TA's wbsite) the system crashes with assetion failures in the manager or router (????). Anyone has been able to get this up and running ? Let me know at shetye@usc.edu.

Thanks !
Siddharth

A Sample of the error I'm getting
===============<Start of Text>================

[mememe@nunki:1] ~/cs551/proja : proja < painput1
Manager::create_manager_tcp
Manager::create_router_processes
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
RealRouter::init
manager.cc:310: failed assertion `stage_ >= 4'
router.cc:193: failed assertion `router.ccsizeof(config) == read(manager_socket_
, &config, sizeof(config))'
:193: failed assertion `sizeof(config) == read(manager_socket_, &config, sizeof(
config))'
router.cc:193: failed
 assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))router.cc'
Abort
:193: failed assertion `sizeof(config) == read(manager_socket_, &config, sizeof(
config))'
[shetye@nunki:1] ~/cs551/proja : router.cc:193: failed assertion `sizeof(config)
 == read(manager_socket_, &config, sizeof(config))'
router.cc:193: failed assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))'
router.cc:193: failed assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))'
router.cc:193: failed assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))'
router.cc:193: failed assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))'
router.cc:193: failed assertion `sizeof(config) == read(manager_socket_, &config
, sizeof(config))'
======================<end of text>=======================


From johnh@ISI.EDU  Mon Apr  7 23:00:45 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 07 Apr 2003 15:00:45 -0700
Subject: [Csci551-talk] project A solution executing ?
In-Reply-To: <1eb2c361eb293e.1eb293e1eb2c36@usc.edu>
Message-ID: <200304072200.h37M0j7V010557@dash.isi.edu>

On Mon, 07 Apr 2003 14:54:41 PDT, siddharth shetye wrote: 
>Hi,
>I just downloaded the solution to project A from the TA's website and compiled it. But when I test it with painput1 thru painput4 (also recently downloaded from the TA's wbsite) the system crashes with assetion failures in the manager or router (????). Anyone has been able to get this up and running ? Let me know at shetye@usc.edu.
>

I believe it was previously mentioned on csci551-talk and/or my or the
TAs web page that painput[1-4] were not all strictly within the
assignment specification.  As a result, new sample inputs (and
outputs) were posted more than a week before Proj A was due.

If you try the sample code on those (revised) inputs, it should run
without error.

   -John Heidemann


From shetye@usc.edu  Tue Apr  8 00:17:00 2003
From: shetye@usc.edu (siddharth shetye)
Date: Mon, 07 Apr 2003 16:17:00 -0700
Subject: [Csci551-talk] project A solution executing ?
Message-ID: <1f46b581f4a3fe.1f4a3fe1f46b58@usc.edu>

Tried that and it works ! Thanks a lot !! I had forgotten that the older files were out of spec ... sorry for that.

- Siddharth

----- Original Message -----
From: John Heidemann <johnh@isi.edu>
Date: Monday, April 7, 2003 3:00 pm
Subject: Re: [Csci551-talk] project A solution executing ?

> On Mon, 07 Apr 2003 14:54:41 PDT, siddharth shetye wrote: 
> >Hi,
> >I just downloaded the solution to project A from the TA's website and 
> compiled it. But when I test it with painput1 thru painput4 (also recently 
> downloaded from the TA's wbsite) the system crashes with assetion failures 
> in the manager or router (????). Anyone has been able to get this up and 
> running ? Let me know at shetye@usc.edu.
> >
> 
> I believe it was previously mentioned on csci551-talk and/or my or the
> TAs web page that painput[1-4] were not all strictly within the
> assignment specification.  As a result, new sample inputs (and
> outputs) were posted more than a week before Proj A was due.
> 
> If you try the sample code on those (revised) inputs, it should run
> without error.
> 
>   -John Heidemann


From gnawali@usc.edu  Tue Apr  8 08:41:33 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Tue, 08 Apr 2003 00:41:33 -0700
Subject: [Csci551-talk] Trace format
Message-ID: <200304080741.h387fXfF005647@enl.usc.edu>

The link on my page to wireless trace information was wrong. Now it
correctly points to a page that explains the fields of a trace and
even has an example line and an explanation of that line towards the
bottom of the page.

From sarangar@usc.edu  Wed Apr  9 02:49:45 2003
From: sarangar@usc.edu (Ramesh Sarangarajan)
Date: Tue, 08 Apr 2003 18:49:45 -0700
Subject: [Csci551-talk] Re: [Csci555-chat] Homeworks
In-Reply-To: <20030408212635.GA38936@pun.isi.edu>
Message-ID: <5.1.0.14.2.20030408184726.00b1eb70@email.usc.edu>

IMHO, we need a couple more HWs before facing the finals. We had and still 
have a lot of time for term paper.

-Ramesh

At 02:26 PM 4/8/2003 -0700, you wrote:
>I haven't been assigning homeworks lately because I figured you were all
>concentrating on your papers and proposals.  However I recognize that
>homeworks may help you all prepare for the final or get a couple more
>grades on record.  It's tough for me to get a feel for the tradeoffs to
>you.  I have no desire to assign busy work, especially when you all have
>the daunting and hopefully exciting prospect of researching an OS area
>and composing a paper about it; however, I don't want to miss helping
>you
>with the upcoming final, either.
>
>So here's what I want to know:  are you interested in having more
>homeworks before the final?
>
>Feel free to reply either to the list at large or to me.
>
>--
>----------------------------------------------------------------------
>Ted Faber                                                faber@isi.edu
>USC/ISI Computer Scientist                   http://www.isi.edu/~faber
>(310) 448-9190         PGP Keys: http://www.isi.edu/~faber/pubkeys.asc


From sarangar@usc.edu  Wed Apr  9 02:51:18 2003
From: sarangar@usc.edu (Ramesh Sarangarajan)
Date: Tue, 08 Apr 2003 18:51:18 -0700
Subject: [Csci551-talk] sorry for posting in the wrong list
Message-ID: <5.1.0.14.2.20030408185007.00b2f1a0@email.usc.edu>

I'm very sorry for posting my previous message in 551 list instead of 555 list.

-Ramesh


From sbhavsar@usc.edu  Wed Apr  9 09:35:13 2003
From: sbhavsar@usc.edu (siddharth bhavsar)
Date: Wed, 09 Apr 2003 01:35:13 -0700
Subject: [Csci551-talk] MACAW or 802.11 ?
Message-ID: <2759fc32755820.27558202759fc3@usc.edu>

Respected Sir ,
In the HW2 Q2 , are the expected answers to be as per MACAW or 802.11 like for example the expected packet exchanges ?
Thanks and Regards
Siddharth

----- Original Message -----
From: Omprakash Gnawali <gnawali@usc.edu>
Date: Tuesday, April 8, 2003 0:41 am
Subject: [Csci551-talk] Trace format

> 
> The link on my page to wireless trace information was wrong. Now it
> correctly points to a page that explains the fields of a trace and
> even has an example line and an explanation of that line towards the
> bottom of the page.
> 


From mordani@aludra.usc.edu  Wed Apr  9 09:36:32 2003
From: mordani@aludra.usc.edu (Rohit Mordani)
Date: Wed, 09 Apr 2003 01:36:32 -0700 (PDT)
Subject: [Csci551-talk] Trace file format help for ns
Message-ID: <Pine.GSO.4.33.0304090128020.26128-100000@aludra.usc.edu>

Hi this is regarding the trace file format(Wireless) for ns. When the
simulation starts up the nodes send a "message" before doing any RTS-CTS
or data transfer. What is this "message"? This is shown below..(from the
tracefile)
s 0.001216201 _0_ MAC  --- 0 message 84 [0 ffffffff 0 800] ------- [0:255
-1:255 32 0]
r 0.001888684 _1_ MAC  --- 0 message 32 [0 ffffffff 0 800] ------- [0:255
-1:255 32 0]

Another question is that the Agent(cbr) sends a message  before any
RTS-CTS. How is this possible? what is this message? This is also shown
below frm the tracefile
s 1.000000000 _0_ AGT  --- 1 cbr 1000 [0 0 0 0] ------- [0:0 1:0 32 0] [0]
0 0

I didnt understand this line too from the tracefile
D 1.107142857 _0_ RTR  IFQ 2 cbr 1020 [0 0 0 0] ------- [0:0 1:0 32 0] [1]
0 0


Could someone please give me some insight to what to do. I have seen the
link to the Wireless Trace page....but that didnt help much to understand
what these packets are...Even the cmu-trace.cc and cmu-trace.h file didnt
help much.
Thanks
Rohit

	---- R O H I T   M O R D A N I ----
		 rohit.mordani.com


From gurtu@usc.edu  Wed Apr  9 10:40:00 2003
From: gurtu@usc.edu (anurag gurtu)
Date: Wed, 09 Apr 2003 02:40:00 -0700
Subject: [Csci551-talk] Question 2 part e
Message-ID: <276c3f02768ebe.2768ebe276c3f0@usc.edu>

A doubt in Q2e plz clarify. e) What will happen if left and center are sending and right tries to send at the same time?

Whome are left and centre sending too. Eachother or to right?


From xiw@usc.edu  Wed Apr  9 20:45:38 2003
From: xiw@usc.edu (xi wang)
Date: Wed, 09 Apr 2003 12:45:38 -0700
Subject: [Csci551-talk] Trace file format help for ns
Message-ID: <98d17964c7.964c798d17@usc.edu>

The AGT cbr seems to indicate a packet has arrived from a higher lever inside a node. Haven't found document though.

Xi


----- Original Message -----
From: Rohit Mordani <mordani@aludra.usc.edu>
Date: Wednesday, April 9, 2003 1:36 am
Subject: [Csci551-talk] Trace file format help for ns

> Hi this is regarding the trace file format(Wireless) for ns. When the
> simulation starts up the nodes send a "message" before doing any RTS-CTS
> or data transfer. What is this "message"? This is shown below..(from the
> tracefile)
> s 0.001216201 _0_ MAC  --- 0 message 84 [0 ffffffff 0 800] ------- [0:255
> -1:255 32 0]
> r 0.001888684 _1_ MAC  --- 0 message 32 [0 ffffffff 0 800] ------- [0:255
> -1:255 32 0]
> 
> Another question is that the Agent(cbr) sends a message  before any
> RTS-CTS. How is this possible? what is this message? This is also shown
> below frm the tracefile
> s 1.000000000 _0_ AGT  --- 1 cbr 1000 [0 0 0 0] ------- [0:0 1:0 32 0] [0]
> 0 0
> 
> I didnt understand this line too from the tracefile
> D 1.107142857 _0_ RTR  IFQ 2 cbr 1020 [0 0 0 0] ------- [0:0 1:0 32 0] [1]
> 0 0
> 
> 
> Could someone please give me some insight to what to do. I have seen the
> link to the Wireless Trace page....but that didnt help much to understand
> what these packets are...Even the cmu-trace.cc and cmu-trace.h file didnt
> help much.
> Thanks
> Rohit
> 
> 	---- R O H I T   M O R D A N I ----
>                 rohit.mordani.com
> 
> 


From akulkarn@usc.edu  Wed Apr  9 21:13:23 2003
From: akulkarn@usc.edu (anupam kulkarni)
Date: Wed, 09 Apr 2003 13:13:23 -0700
Subject: [Csci551-talk] tracefile
Message-ID: <c3e61c35f1.c35f1c3e61@usc.edu>

hi
i think the 1st message tht u r talking abt is the DSDV msg. so node 0 sends a DSDV msg .node 1 receives it. and by the time node 1 sends a DSDV msg back to 0 the AGT of node 0 has made a few packets. 
I would like to know more abt the trace file myself.
for example the queue length is some 36K something. but still the 1st AGT packet is dropped. anyone having info plz let us know
Anupam



From gnawali@usc.edu  Thu Apr 10 04:27:48 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 20:27:48 -0700
Subject: [Csci551-talk] trace file format
Message-ID: <200304100327.h3A3Rmw6002800@enl.usc.edu>

Dear class,

I have posted another link on my website that will help you understand
the trace file format.

- om_p

From gnawali@usc.edu  Thu Apr 10 04:37:10 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 20:37:10 -0700
Subject: [Csci551-talk] hw2 q2
Message-ID: <200304100337.h3A3bAX1002922@enl.usc.edu>

In the second question, When you try out case(d) with the simulator
using the script, you need to make sure the nodes are equidistant
before you simulate. As one student pointed out, they are not
equidistant to being with.

From gnawali@usc.edu  Thu Apr 10 05:01:37 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 21:01:37 -0700
Subject: [Csci551-talk] hw2 q1
Message-ID: <200304100401.h3A41bix003467@enl.usc.edu>

Dear class,

For, hw2 q1, parts (b) and (c), you should provide mathematical
expresion.

- om_p

From gnawali@usc.edu  Thu Apr 10 05:14:41 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 21:14:41 -0700
Subject: [Csci551-talk] hw2 q2
Message-ID: <200304100414.h3A4Efgd003637@enl.usc.edu>

Some students wanted clarification as to which node is sending data to
which other node in the network (second part of the second question of
the second homework). Here is a hint: If you think about how wireless
communication works, it should be enough to know that they are
"sending" to answer the 2e

From gnawali@usc.edu  Thu Apr 10 05:26:08 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 21:26:08 -0700
Subject: [Csci551-talk] MACAW or 802.11 ?
In-Reply-To: Your message of Wed, 09 Apr 2003 01:35:13 -0700
Message-ID: <200304100426.h3A4Q8C3003801@enl.usc.edu>

You can answer thought questions based on MACAW. The simulations use
802.11.

you wrote:
> Respected Sir ,
> In the HW2 Q2 , are the expected answers to be as per MACAW or 802.11 like fo
 *r example the expected packet exchanges ?
> Thanks and Regards
> Siddharth
> 
> ----- Original Message -----
> From: Omprakash Gnawali <gnawali@usc.edu>
> Date: Tuesday, April 8, 2003 0:41 am
> Subject: [Csci551-talk] Trace format
> 
> > 
> > The link on my page to wireless trace information was wrong. Now it
> > correctly points to a page that explains the fields of a trace and
> > even has an example line and an explanation of that line towards the
> > bottom of the page.
> > 

From gnawali@usc.edu  Thu Apr 10 05:28:05 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 09 Apr 2003 21:28:05 -0700
Subject: [Csci551-talk] Trace file format help for ns
In-Reply-To: Your message of Wed, 09 Apr 2003 01:36:32 -0700 (PDT)
Message-ID: <200304100428.h3A4S5eG003825@enl.usc.edu>

The agent sends the message to the lower layers. The lower layer will
initiate RTS/CTS exchanges when it has to data to transfer.

> 
> Another question is that the Agent(cbr) sends a message  before any
> RTS-CTS. How is this possible? what is this message? This is also shown
> below frm the tracefile
> s 1.000000000 _0_ AGT  --- 1 cbr 1000 [0 0 0 0] ------- [0:0 1:0 32 0] [0]
> 0 0
> 

From dwipalde@usc.edu  Fri Apr 11 00:49:05 2003
From: dwipalde@usc.edu (Dwipal Desai)
Date: Thu, 10 Apr 2003 16:49:05 -0700
Subject: [Csci551-talk] PDF files not downloadable on class website..
In-Reply-To: <98d17964c7.964c798d17@usc.edu>
References: <98d17964c7.964c798d17@usc.edu>
Message-ID: <1050018545.2453.9.camel@changsha>

Hi

I am not able to download the papers from class website.

Anyone faced the same problem ? 

-Dwipal


From gnawali@usc.edu  Fri Apr 11 01:11:24 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Thu, 10 Apr 2003 17:11:24 -0700
Subject: [Csci551-talk] PDF files not downloadable on class website..
In-Reply-To: Your message of Thu, 10 Apr 2003 16:49:05 -0700
Message-ID: <200304110011.h3B0BOcc014004@enl.usc.edu>

Thanks for letting us know. We will fix this soon. If you want to read
the papers right away, what you can do is go to Google and search for
the title of the paper. You will see online copies of the paper.

you wrote:
> Hi
> 
> I am not able to download the papers from class website.
> 
> Anyone faced the same problem ? 
> 
> -Dwipal

From shshanks@usc.edu  Fri Apr 11 01:08:17 2003
From: shshanks@usc.edu (Shshank Sharma)
Date: Thu, 10 Apr 2003 17:08:17 -0700
Subject: [Csci551-talk] PDF files not downloadable on class website..
In-Reply-To: <1050018545.2453.9.camel@changsha>
Message-ID: <HKEPLAIFHFMJHEAMANJCEENACBAA.shshanks@usc.edu>

> I am not able to download the papers from class website.
>Anyone faced the same problem ? 

Yes, me.

From johnh@ISI.EDU  Fri Apr 11 01:11:49 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Thu, 10 Apr 2003 17:11:49 -0700
Subject: [Csci551-talk] PDF files not downloadable on class website..
In-Reply-To: <1050018545.2453.9.camel@changsha>
Message-ID: <200304110011.h3B0Bnkf005662@dash.isi.edu>

On Thu, 10 Apr 2003 16:49:05 PDT, Dwipal Desai wrote: 
>Hi
>
>I am not able to download the papers from class website.
>
>Anyone faced the same problem ? 
>
>-Dwipal

This problem should now be fixed.

   -John Heidemann

From shetye@usc.edu  Fri Apr 11 09:11:29 2003
From: shetye@usc.edu (Siddharth Shetye)
Date: Fri, 11 Apr 2003 01:11:29 -0700
Subject: [Csci551-talk] DoD's next generation network
References: <200304101905.h3AJ56k14857@gamma.isi.edu>
Message-ID: <000701c30001$f4e743b0$06107e18@darkstar>

Slightly unrelated to CS551 if we were to follow the syllabus as listed on
the course website, but I thought people taking 551 could be interested in
this. Some networking related development. If you are too bogged down by
papers/project/hw, forget this email and go on !

The DoD just unveiled plans for a next generation network to be used by
their military in air, water/underwater, on land and in space. I tried
digging up some details about this and found this one below ... don't know
any specific details ....
http://www.commsdesign.com/story/OEG20030409S0034

Nuclear survivability sounds good. So is support for thin-client
(interactive thin clients across the globe?? ... wonder what tricks they are
using!).

Siddharth


From johnh@ISI.EDU  Fri Apr 11 20:44:59 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Fri, 11 Apr 2003 12:44:59 -0700
Subject: [Csci551-talk] DoD's next generation network
In-Reply-To: <000701c30001$f4e743b0$06107e18@darkstar>
Message-ID: <200304111944.h3BJix0O006496@dash.isi.edu>

On Fri, 11 Apr 2003 01:11:29 PDT, "Siddharth Shetye" wrote: 
>Slightly unrelated to CS551 if we were to follow the syllabus as listed on
>the course website, but I thought people taking 551 could be interested in
>this. Some networking related development. If you are too bogged down by
>papers/project/hw, forget this email and go on !
>
>The DoD just unveiled plans for a next generation network to be used by
>their military in air, water/underwater, on land and in space. I tried
>digging up some details about this and found this one below ... don't know
>any specific details ....
>http://www.commsdesign.com/story/OEG20030409S0034
>
>Nuclear survivability sounds good. So is support for thin-client
>(interactive thin clients across the globe?? ... wonder what tricks they are
>using!).

At the level of that write up this could be anything.
But switching circuit -> packet sounds like it's looking at moving
their voice traffic to all packet data, like some telcos have
proposed (but not carried out).

   -John Heidemann


From johnh@ISI.EDU  Fri Apr 11 21:20:04 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Fri, 11 Apr 2003 13:20:04 -0700
Subject: [Csci551-talk] MACAW or 802.11 ?
In-Reply-To: <200304100426.h3A4Q8C3003801@enl.usc.edu>
Message-ID: <200304112020.h3BKK4Eb006764@dash.isi.edu>

On Wed, 09 Apr 2003 21:26:08 PDT, Omprakash Gnawali wrote: 
>
>You can answer thought questions based on MACAW. The simulations use
>802.11.

I think Om meant is:
even though the simulations use 802.11,
you can answer the question using your knowledge of MACAW
(as if the simulations ran MACAW)
because, for this question, the answers should not be different
between the two protocols.

   -John Heidemann


From gnawali@usc.edu  Mon Apr 14 21:54:25 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Mon, 14 Apr 2003 13:54:25 -0700
Subject: [Csci551-talk] compiling sample proja code
Message-ID: <200304142054.h3EKsPD2008049@enl.usc.edu>

Some students were still having problem compiling the sample proja
code on Solaris. This was because of non standard values on
environment variables depending on the shell used. I decided to make
it a lot simpler by having the Makefile work by default on Solaris. If
you want to compile the code on Linux, you need to uncomment a line
from the Makefile. There is a comment on the Makefile that tells you
which line to comment to compile on Linux.

If you are still having problems, please let me know.

- om_p

From jvora@usc.edu  Wed Apr 16 05:41:57 2003
From: jvora@usc.edu (jigesh vora)
Date: Tue, 15 Apr 2003 21:41:57 -0700
Subject: [Csci551-talk] Project B
Message-ID: <4b1a034b1fda.4b1fda4b1a03@usc.edu>

Hi,
   According to the specifications for stage 6 the master writes in the file s6-n.out. However it would be straightforward if each router writes into its file. So is there a typo error in the spec???

Thanx
Jigesh 


From shshanks@aludra.usc.edu  Wed Apr 16 14:04:42 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Wed, 16 Apr 2003 06:04:42 -0700 (PDT)
Subject: [Csci551-talk] Project B
In-Reply-To: <4b1a034b1fda.4b1fda4b1a03@usc.edu>
Message-ID: <Pine.GSO.4.33.0304160600090.25486-100000@aludra.usc.edu>

On Tue, 15 Apr 2003, jigesh vora wrote:

> Hi,
>    According to the specifications for stage 6 the master writes in the file s6-n.out.
> However it would be straightforward if each router writes into its file.
> So is there a typo error in the spec???

I also feel each router doing this is more convenient than relaying the
info to the manager every time a router takes some action.
Also Stage 7 output says:
	Each router should create a file "s7-n.out" (where n is the router id).
	This file should log all the output from stage6, plus it should log
	..."
This gives that each router should be doing the logging in Stage 6 also.
Could the TA or the Professor please confirm this ?

Thanks
-Sharma


From shshanks@aludra.usc.edu  Wed Apr 16 14:59:09 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Wed, 16 Apr 2003 06:59:09 -0700 (PDT)
Subject: [Csci551-talk] Project-B: routability
Message-ID: <Pine.GSO.4.33.0304160313470.11286-100000@aludra.usc.edu>

Hi,
This might be pretty basic, but just want to make sure I have understood
it correctly. In deciding routability of a destination-IP-address when
asked by the manager, is the following "all" that a router must do ?

- Check if the corresponding parts in the sought destination-IP-address
and the network part of one of the IP-addresses in its routing table
match. Assume routability if they do and return "no route" if they don't.

Something I am missing ?





From johnh@ISI.EDU  Wed Apr 16 13:40:26 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 16 Apr 2003 05:40:26 -0700
Subject: [Csci551-talk] Project B
In-Reply-To: <4b1a034b1fda.4b1fda4b1a03@usc.edu>
Message-ID: <200304161240.h3GCeQmE002010@dash.isi.edu>

On Tue, 15 Apr 2003 21:41:57 PDT, jigesh vora wrote: 
>Hi,
>   According to the specifications for stage 6 the master writes in the file s6-n.out. However it would be straightforward if each router writes into its file. So is there a typo error in the spec???
>
>Thanx
>Jigesh 

Yes, that is a typo.  The first sentence of section 3 "Output"
section  should read "Each router should create the file s6-n.out ...

Thanks for pointing that out.
   -John Heidemann



From johnh@ISI.EDU  Wed Apr 16 17:24:11 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 16 Apr 2003 09:24:11 -0700
Subject: [Csci551-talk] second bug in proj b
Message-ID: <200304161624.h3GGOBUS003097@dash.isi.edu>

The sample input in section 4 / stage7 of Project B is incorrect:

the last two join requests don't include RP addresses

Here is a corrected sample input:


# mcast establishment and sending:
# action initiating-router-id multicast-group-address
j 1 10.1.0.1    224.0.10.1
s 1 224.0.10.1
s 1 224.0.10.1
j 1 192.168.101.15    225.0.101.15
s 1 225.0.101.15
s 3 225.0.101.15
s 1 224.0.10.1
j 5 192.168.123.255 225.0.123.255
s 5 225.0.123.255
j 1 11.0.1.1    2224.0.11.1
j 3 192.168.101.14 255.0.101.14
j 4 192.168.101.15 225.0.101.15
s 1 225.0.101.15
j 3 192.168.101.15    225.0.101.15
s 1 225.0.101.15
j 0 0.0.0.0 0.0.0.0


As with Project A, you may assume that the input is syntactically
correct.  (It is not your responsibility to deal with inputs with
joins that omit RPs.)

   -John Heidemann



From johnh@ISI.EDU  Wed Apr 16 14:23:18 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 16 Apr 2003 06:23:18 -0700
Subject: [Csci551-talk] project B corrections
Message-ID: <200304161323.h3GDNIFm002327@dash.isi.edu>

Two bugs in the sample output for Project B:

in section 3
the first line of the "output" section should be "Each router will
create..."
not "The master will create..."

section 4 sample output, routes to 225.0.101.15 shoudl go to 2, not 3.
i.e.:

the line
sending packet id 3 group 225.0.101.15 to 3

should be
sending packet id 3 group 225.0.101.15 to 2



   -John Heidemann

From gnawali@usc.edu  Thu Apr 17 20:51:19 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Thu, 17 Apr 2003 12:51:19 -0700
Subject: [Csci551-talk] sample test for project A
Message-ID: <200304171951.h3HJpJtN017311@enl.usc.edu>

Some students wanted test cases we used to grade project A. I have
posted them on my website. This would be a good way of finding out if
your project A implementation is functioning reasonably.

From shshanks@aludra.usc.edu  Thu Apr 17 22:19:18 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Thu, 17 Apr 2003 14:19:18 -0700 (PDT)
Subject: [Csci551-talk] sample test for project A
In-Reply-To: <200304171951.h3HJpJtN017311@enl.usc.edu>
Message-ID: <Pine.GSO.4.33.0304171413230.24126-100000@aludra.usc.edu>

The assumption that I (and many people seem to have followed) is that the
last line of an input file for project A:
0	0	0.0.0.0		0 0

will end with a '\n'.
This is not true for any of the input files.

Consequently, segfaults abound.
I made that change in the files, the file was read correctly, there
was no ugly segfault, routers started talking, the tables started
getting written into and the world was a happy place once more.
But not quite, yet.

-Sharma

___________________________________________

On Thu, 17 Apr 2003, Omprakash Gnawali wrote:

>
> Some students wanted test cases we used to grade project A. I have
> posted them on my website. This would be a good way of finding out if
> your project A implementation is functioning reasonably.
>


From johnh@ISI.EDU  Thu Apr 17 23:59:51 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Thu, 17 Apr 2003 15:59:51 -0700
Subject: [Csci551-talk] sample test for project A
In-Reply-To: <200304171951.h3HJpJtN017311@enl.usc.edu>
Message-ID: <200304172300.h3HMxpns003053@dash.isi.edu>

On Thu, 17 Apr 2003 12:51:19 PDT, Omprakash Gnawali wrote: 
>
>Some students wanted test cases we used to grade project A. I have
>posted them on my website. This would be a good way of finding out if
>your project A implementation is functioning reasonably.

Although, the Project A test cases are NOT guaranteed to offer
complete coverage.  Students should still think about different cases
the algorithms raise and design your own test cases.  (For Project B.)

   -John Heidemann
 

From gurtu@usc.edu  Fri Apr 18 07:46:06 2003
From: gurtu@usc.edu (anurag gurtu)
Date: Thu, 17 Apr 2003 23:46:06 -0700
Subject: [Csci551-talk] stage 6 output file
Message-ID: <fa8f6ffa9633.fa9633fa8f6f@usc.edu>

Greetings, 
1)In stage6: Is output of s6-3.out correct.
                  joining 225.0.101.15 due to request from 1
shoudnt it be joining 255.0.101.15 due to request from 3 , instead.

2) Also will it be possible to give us some sample input test cases for stage 6and 7.

Thanks
Anurag Gurtu  



From johnh@ISI.EDU  Fri Apr 18 16:25:22 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Fri, 18 Apr 2003 08:25:22 -0700
Subject: [Csci551-talk] stage 6 output file
In-Reply-To: <fa8f6ffa9633.fa9633fa8f6f@usc.edu>
Message-ID: <200304181525.h3IFPMrI005519@dash.isi.edu>

On Thu, 17 Apr 2003 23:46:06 PDT, anurag gurtu wrote: 
>Greetings, 
>1)In stage6: Is output of s6-3.out correct.
>                  joining 225.0.101.15 due to request from 1
>shoudnt it be joining 255.0.101.15 due to request from 3 , instead.

You are correct (the original handout was wrong).
I already posted about the incorrect routing for 225.0.101.15 on
csci551-talk (and it should be on Om's web page as well).

>2) Also will it be possible to give us some sample input test cases for stage 6and 7.

Beyond what's in the handout?  No.  You should make your own
additional test cases.

   -John Heidemann

From phadke@usc.edu  Fri Apr 18 20:15:54 2003
From: phadke@usc.edu (amol-uday phadke)
Date: Fri, 18 Apr 2003 12:15:54 -0700
Subject: [Csci551-talk] Re: Csci551-talk digest, Vol 1 #132 - 5 msgs
Message-ID: <115309f114fcfc.114fcfc115309f@usc.edu>

Hi
Many students have questions regarding the last dv-message and the common questions which seem to arise are
If I get no points for this part then how is my code is working,
Here is an explanation
The 5 points for this part were for the correct marshalling of the bytes when you send out the message.
The dv message should have the routes field
same but reordered, and the number of routes
field should denote the total routes ( at 
position 15-18) for dv-1 and at 11-14 for 
dv-0. If you havent marshalled data correctly, I wont find what I am looking for at the exact positions, so though your code is working because the message is passed correctly and parsed correctly, from the network point of view marshalling hasnt happened and hence the points will be taken off.
What I would suggest is please have a look at the sample OUTPUTS which we posted on the web for projectA and compare it what you got. You need to compare the exacxt positions. As an example the sample output follows dv-1 message and hence positions 15-18 will have 000f for 15 routes in the final routing table.
I hope this answers doubts regarding the dv-message
Thanks
..amol. 



From johnh@ISI.EDU  Fri Apr 18 22:55:21 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Fri, 18 Apr 2003 14:55:21 -0700
Subject: [Csci551-talk] a couple more proj b clarifications
Message-ID: <200304182155.h3ILtLP2007176@dash.isi.edu>

1. In Section 3, change the statement
"...or if the multicast address is not in the IP address range"
to 
"...or if the multicast address is not in the IP multicast address range"
[Add multicast]

(You can assume that all addresses will be valid IP addresses.)



2. After talking with Amol about Project A, the sample test cases he
posted were actually a *simplified* form of the test cases that we
expected to be used.

The complete Project A test cases will be posted to Om's web page
shortly.  All students are advised to check that their Project A code
works on it and that their Project B code works on similar input.

(There will be newlines at the end of the last line of any Project B
test cases.)

   -John Heidemann

From johnh@ISI.EDU  Sat Apr 19 00:19:59 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Fri, 18 Apr 2003 16:19:59 -0700
Subject: [Csci551-talk] defining "joining" in project B
Message-ID: <200304182320.h3INJxDZ002329@dash.isi.edu>

A question that came up in class today about project b:

When nodes log if they join the mcast group,
they should record this both if they join because of a local
client, OR if they join because someone downstream requires them to
forward packets.

I.e., in the topology 

E---A---B---C---D

If a user at A joins a group G with an RP at C
A would log   "joining G due to request from A"
B would log   "joining G due to request from A"
C would log   "joining G due to request from B"
(and D wouldn't log anything).

(I.e., B is a joiner even though it just transits traffic.)

Something to think about: if E then joins G, who needs to log?

   -John Heidemann


From rajgarhi@usc.edu  Sat Apr 19 07:03:58 2003
From: rajgarhi@usc.edu (abhishek rajgarhia)
Date: Fri, 18 Apr 2003 23:03:58 -0700
Subject: [Csci551-talk] Project B Query
Message-ID: <138889d1389116.1389116138889d@usc.edu>

Hi,

May we assume that the address of all the RP's will fall under the domain of some internal router, since we are not having processes for any external routers.

Thank you

Regards

Abhishek 


From mahuja@usc.edu  Sat Apr 19 13:17:20 2003
From: mahuja@usc.edu (mona ahuja)
Date: Sat, 19 Apr 2003 05:17:20 -0700
Subject: [Csci551-talk] Question regarding stage 7 input
Message-ID: <13c93ea13c9feb.13c9feb13c93ea@usc.edu>

Professor,
The project specification says
that " You may assume that all nodes that will send data are on the multicast tree"

I feel that the input for stage 7 may have an error . 
line 6  s 3 225.0.101.15
should be 
s 2 225.0.101.15 as 3 is not yet a part of multicast tree.
the tree for multicast group contains 1-2-4 
PLease correct me if I am wrong.
ThankYou
Mona


From mahuja@usc.edu  Sun Apr 20 02:07:15 2003
From: mahuja@usc.edu (mona ahuja)
Date: Sat, 19 Apr 2003 18:07:15 -0700
Subject: [Csci551-talk] proj b doubt
Message-ID: <14de43314dfa98.14dfa9814de433@usc.edu>

i have doubt
Can we safely assume that no two routers in the input file will have same addr 
mask . 
If we cannot make this assumption then do  we have to use some prioritisation rules to deicide which router to choose as the next hop .
Thank you
Mona




From johnh@ISI.EDU  Sun Apr 20 02:32:20 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sat, 19 Apr 2003 18:32:20 -0700
Subject: [Csci551-talk] Project B Query
In-Reply-To: <138889d1389116.1389116138889d@usc.edu>
Message-ID: <200304200132.h3K1WKbn032204@dash.isi.edu>

On Fri, 18 Apr 2003 23:03:58 PDT, abhishek rajgarhia wrote: 
>Hi,
>
>May we assume that the address of all the RP's will fall under the domain of some internal router, since we are not having processes for any external routers.
>
>Thank you
>
>Regards
>
>Abhishek 


There will be no RP addresses routed by external routers.
They will all be either internal routers or unrouted address space.

   -John Heidemann

From johnh@ISI.EDU  Sun Apr 20 02:39:06 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sat, 19 Apr 2003 18:39:06 -0700
Subject: [Csci551-talk] Re: Question regarding stage 7 input
In-Reply-To: <13c93ea13c9feb.13c9feb13c93ea@usc.edu>
Message-ID: <200304200139.h3K1d6Y4032266@dash.isi.edu>

On Sat, 19 Apr 2003 05:17:20 PDT, mona ahuja wrote: 
>
>Professor,
>The project specification says
>that " You may assume that all nodes that will send data are on the multicast tree"
>
>I feel that the input for stage 7 may have an error . 
>line 6  s 3 225.0.101.15
>should be 
>s 2 225.0.101.15 as 3 is not yet a part of multicast tree.
>the tree for multicast group contains 1-2-4 
>PLease correct me if I am wrong.
>ThankYou
>Mona

You are correct.  I had previously posted that the original handout
had incorrectly said that 1 will route multicast group 225.0.101.15
through 2, not 3 (as it originally said).  This line needs to change
due to that.

   -John Heidemann

From johnh@ISI.EDU  Sun Apr 20 02:39:56 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sat, 19 Apr 2003 18:39:56 -0700
Subject: [Csci551-talk] project B clarifications
Message-ID: <200304200139.h3K1du3k032302@dash.isi.edu>

Several additional project B clarifications:

1. You do not need to do full longest prefix matching.  While the
masks will vary in size and must be considered, we will not give
overlapping addresses.

2. There will be no RP addresses routed by external routers.
They will all be either internal routers or unrouted address space.

3. In stage 7, line 6 "s 3 225.0.101.15" should be "s 2 225.0.101.15",
consistent with the previous changes that said all references to group
225.0.101.15 and node 3 should use node 2.



   -John Heidemann

From jvora@usc.edu  Sun Apr 20 18:06:48 2003
From: jvora@usc.edu (jigesh vora)
Date: Sun, 20 Apr 2003 10:06:48 -0700
Subject: [Csci551-talk] Project B: Reliable communication over UDP
Message-ID: <15be3df15ba49d.15ba49d15be3df@usc.edu>

Hi,
  I just wanted to confirm that the lossrate for stages 6 and 7 will be same as the lossrate given for the previous stages in Project A and there will be packet drops over UDP depending on that loss rate.Please correct me if I am wrong? 

Thanx
Jigesh




From gnawali@usc.edu  Sun Apr 20 18:36:37 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Sun, 20 Apr 2003 10:36:37 -0700
Subject: [Csci551-talk] Project B: Reliable communication over UDP
In-Reply-To: Your message of Sun, 20 Apr 2003 10:06:48 -0700
Message-ID: <200304201736.h3KHac0a030656@enl.usc.edu>

Correct.

you wrote:
> Hi,
>   I just wanted to confirm that the lossrate for stages 6 and 7 will be same as t
 *he lossrate given for the previous stages in Project A and there will be packet d
 *rops over UDP depending on that loss rate.Please correct me if I am wrong? 
> 
> Thanx
> Jigesh
> 
> 

From shshanks@aludra.usc.edu  Sun Apr 20 18:44:01 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Sun, 20 Apr 2003 10:44:01 -0700 (PDT)
Subject: [Csci551-talk] Project B: Reliable communication over UDP
In-Reply-To: <15be3df15ba49d.15ba49d15be3df@usc.edu>
Message-ID: <Pine.GSO.4.33.0304201037140.15349-100000@aludra.usc.edu>


> Hi,
>   I just wanted to confirm that the lossrate for stages
> 6 and 7 will be same as the lossrate given for the previous stages in
> Project A and there will be packet drops over UDP depending on that loss
> rate.Please correct me if I am wrong?

Further to this, what max loss rate would the code be checked against,
since the timer for convergence (wrt project-A) does depend on this value,
more severely for long linear topologies with high loss rate.
Having a topology-aware adaptive _convergence_ timer would be good, but we
can perhaps make do with a simple assumption here ?

-Sharma


From jvora@usc.edu  Sun Apr 20 19:01:11 2003
From: jvora@usc.edu (jigesh vora)
Date: Sun, 20 Apr 2003 11:01:11 -0700
Subject: [Csci551-talk] TTL in project B
Message-ID: <15cd06b15d2551.15d255115cd06b@usc.edu>

Hi,
  since we are using only muticast trees for communication there will not be any loops. So can someone tell me the use of TTL in the messages?

Thanx
Jigesh


From xiw@usc.edu  Sun Apr 20 19:15:13 2003
From: xiw@usc.edu (Xi Wang)
Date: Sun, 20 Apr 2003 11:15:13 -0700
Subject: [Csci551-talk] Drop data packet
References: <15cd06b15d2551.15d255115cd06b@usc.edu>
Message-ID: <001e01c30768$ca49e720$efe57d80@xwl3>

Hi. My understanding about packet loss is that we also need to drop data
packets (s command) without retransmission. Is this correct?

Thanks,

Xi


From shshanks@aludra.usc.edu  Sun Apr 20 19:42:50 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Sun, 20 Apr 2003 11:42:50 -0700 (PDT)
Subject: [Csci551-talk] Drop data packet
In-Reply-To: <001e01c30768$ca49e720$efe57d80@xwl3>
Message-ID: <Pine.GSO.4.33.0304201141401.15349-100000@aludra.usc.edu>

> Hi. My understanding about packet loss is that we also need to drop data
> packets (s command) without retransmission. Is this correct?
>

Methinks, those and the Joins have to be "sent reliably" as the
assigment says. All UDP has to be made reliable.

-Sharma


From shetye@usc.edu  Mon Apr 21 02:22:37 2003
From: shetye@usc.edu (Siddharth Shetye)
Date: Sun, 20 Apr 2003 18:22:37 -0700
Subject: [Csci551-talk] Join logging
References: <200304191905.h3JJ51028091@gamma.isi.edu>
Message-ID: <000e01c307a4$7e6ab820$6401a8c0@darkstar>

Hi,

Just a minor question about logging ...

Do we have to log the join messages only when the join request is for a
"new" RP/Multicast Group or do we also have to log the "Joining a.b.c.d due
to request from xyz" information even when the branch is already set in that
node ? In the first case we'd have to update our state and propogate the
information to the next hop while in the second case we only need to update
our state to reflect the new member.

So asking it another way, do we log for 1) every new member or 2) just for
every new mcast group/RP ?

Thanks
Siddharth


From rajgarhi@usc.edu  Mon Apr 21 03:36:27 2003
From: rajgarhi@usc.edu (abhishek rajgarhia)
Date: Sun, 20 Apr 2003 19:36:27 -0700
Subject: [Csci551-talk] Order of route queries
Message-ID: <170614f1702b98.1702b98170614f@usc.edu>

Hi,

Is the order in which the answer route queries are logged in s5.out, important? Does the order have to match the order of the queries in the input file?

Thanks

Abhishek 


From shetye@usc.edu  Mon Apr 21 03:53:19 2003
From: shetye@usc.edu (Siddharth Shetye)
Date: Sun, 20 Apr 2003 19:53:19 -0700
Subject: [Csci551-talk] A question about STL List of pointers ...
References: <200304201905.h3KJ51013697@gamma.isi.edu>
Message-ID: <001401c307b1$2a2a31c0$6401a8c0@darkstar>

Hi,

I was wondering ... if I have the following ...

list<myclass*> list_of_my_class;

it would be a list of pointer to type "myclass" ... so would I be required
to code additional functions into "myclass" (as required by STL objects ,
http://www.msoe.edu/eecs/ce/courseinfo/stl/objreq.htm) ??

I was thinking that since they are a list of only pointers, c++/STL should
be able to handle them natively (and maybe increment etc on the basis of
sizeof(myclass) (or similar) )

Thanks !
Siddharth


From abhished@usc.edu  Mon Apr 21 04:29:39 2003
From: abhished@usc.edu (abhishek dubey)
Date: Sun, 20 Apr 2003 20:29:39 -0700
Subject: [Csci551-talk] A question about STL List of pointers ...
Message-ID: <172d75217263e5.17263e5172d752@usc.edu>

additional functions may or may not be required depending on what operations you intend to perform.
if your requirements are as simple as incrementing, then the answer is no.
you can define an iterator ..
list<myclass>::iterator iterator_name;

simply use iterator_name++ to point to the next element in the list.

abhishek



>Hi,

>I was wondering ... if I have the following ...

>list<myclass*> list_of_my_class;

>it would be a list of pointer to type "myclass" ... so would I be >required
>to code additional functions into "myclass" (as required by STL >objects ,
>http://www.msoe.edu/eecs/ce/courseinfo/stl/objreq.htm) ??

>I was thinking that since they are a list of only pointers, c++/STL >should
>be able to handle them natively (and maybe increment etc on the >basis .of
>sizeof(myclass) (or similar) )

>Thanks !
>Siddharth


From sbhavsar@usc.edu  Mon Apr 21 04:35:20 2003
From: sbhavsar@usc.edu (siddharth bhavsar)
Date: Sun, 20 Apr 2003 20:35:20 -0700
Subject: [Csci551-talk] A question about STL List of pointers ...
Message-ID: <17313261730d22.1730d221731326@usc.edu>

moreover u do not even need the iterator ..
u can as well do with
list<myclass> lst ;
for(..)
lst[i] ; <---
Siddharth

----- Original Message -----
From: abhishek dubey <abhished@usc.edu>
Date: Sunday, April 20, 2003 8:29 pm
Subject: Re:[Csci551-talk] A question about STL List of pointers ...

> additional functions may or may not be required depending on what 
> operations you intend to perform.
> if your requirements are as simple as incrementing, then the answer is no.
> you can define an iterator ..
> list<myclass>::iterator iterator_name;
> 
> simply use iterator_name++ to point to the next element in the list.
> 
> abhishek
> 
> 
> 
> >Hi,
> 
> >I was wondering ... if I have the following ...
> 
> >list<myclass*> list_of_my_class;
> 
> >it would be a list of pointer to type "myclass" ... so would I be >required
> >to code additional functions into "myclass" (as required by STL >objects ,
> >http://www.msoe.edu/eecs/ce/courseinfo/stl/objreq.htm) ??
> 
> >I was thinking that since they are a list of only pointers, c++/STL >should
> >be able to handle them natively (and maybe increment etc on the >basis .of
> >sizeof(myclass) (or similar) )
> 
> >Thanks !
> >Siddharth
> 
> 


From johnh@ISI.EDU  Mon Apr 21 05:37:09 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:37:09 -0700
Subject: [Csci551-talk] proj b doubt
In-Reply-To: <14de43314dfa98.14dfa9814de433@usc.edu>
Message-ID: <200304210437.h3L4b92O018596@dash.isi.edu>

On Sat, 19 Apr 2003 18:07:15 PDT, mona ahuja wrote: 
>i have doubt
>Can we safely assume that no two routers in the input file will have same addr 
>mask . 
>If we cannot make this assumption then do  we have to use some prioritisation rules to deicide which router to choose as the next hop .
>Thank you
>Mona
>
>

I believe Project A  was pretty specific about prioritization rules.

It would be illegal, however, for two different internal routers to
lay claim to the same address space.

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:46:01 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:46:01 -0700
Subject: [Csci551-talk] Project B: Reliable communication over UDP
In-Reply-To: <Pine.GSO.4.33.0304201037140.15349-100000@aludra.usc.edu>
Message-ID: <200304210446.h3L4k1OR018753@dash.isi.edu>

On Sun, 20 Apr 2003 10:44:01 PDT, Shshank Sharma wrote: 
>
>
>> Hi,
>>   I just wanted to confirm that the lossrate for stages
>> 6 and 7 will be same as the lossrate given for the previous stages in
>> Project A and there will be packet drops over UDP depending on that loss
>> rate.Please correct me if I am wrong?
>
>Further to this, what max loss rate would the code be checked against,
>since the timer for convergence (wrt project-A) does depend on this value,
>more severely for long linear topologies with high loss rate.
>Having a topology-aware adaptive _convergence_ timer would be good, but we
>can perhaps make do with a simple assumption here ?
>
>-Sharma

You can assume loss rates <= 25%.

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:49:45 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:49:45 -0700
Subject: [Csci551-talk] TTL in project B
In-Reply-To: <15cd06b15d2551.15d255115cd06b@usc.edu>
Message-ID: <200304210449.h3L4njmr018831@dash.isi.edu>

On Sun, 20 Apr 2003 11:01:11 PDT, jigesh vora wrote: 
>Hi,
>  since we are using only muticast trees for communication there will not be any loops. So can someone tell me the use of TTL in the messages?
>
>Thanx
>Jigesh

You're correct that correct code will not have any loops.

All students in the class who can guarantee that all their code will
work completely correctly the first time are not required to implement
TTLs.

For the rest of the class, you might find that TTLs prevent
catestrophic meltdowns if you accidently get a loop while you're
testing your code.  (Think about it: what happens if you accidently
get a loop where the packets get sent to multiple places... every time
they loop back.)

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:50:52 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:50:52 -0700
Subject: [Csci551-talk] Drop data packet
In-Reply-To: <Pine.GSO.4.33.0304201141401.15349-100000@aludra.usc.edu>
Message-ID: <200304210450.h3L4oqVE018852@dash.isi.edu>

On Sun, 20 Apr 2003 11:42:50 PDT, Shshank Sharma wrote: 
>
>> Hi. My understanding about packet loss is that we also need to drop data
>> packets (s command) without retransmission. Is this correct?
>>
>
>Methinks, those and the Joins have to be "sent reliably" as the
>assigment says. All UDP has to be made reliable.
>
>-Sharma

The spec actually only requires joins to be reliable.

It is your choice whether you send data reliably or not (document your
choice in the README).

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:53:04 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:53:04 -0700
Subject: [Csci551-talk] Join logging
In-Reply-To: <000e01c307a4$7e6ab820$6401a8c0@darkstar>
Message-ID: <200304210453.h3L4r4ev018876@dash.isi.edu>

On Sun, 20 Apr 2003 18:22:37 PDT, "Siddharth Shetye" wrote: 
>Hi,
>
>Just a minor question about logging ...
>
>Do we have to log the join messages only when the join request is for a
>"new" RP/Multicast Group or do we also have to log the "Joining a.b.c.d due
>to request from xyz" information even when the branch is already set in that
>node ? In the first case we'd have to update our state and propogate the
>information to the next hop while in the second case we only need to update
>our state to reflect the new member.
>
>So asking it another way, do we log for 1) every new member or 2) just for
>every new mcast group/RP ?
>
>Thanks
>Siddharth

Every time a router is added to the tree it needs to log that fact.
If it's already on that tree it does not need to log anything.

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:54:21 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:54:21 -0700
Subject: [Csci551-talk] Order of route queries
In-Reply-To: <170614f1702b98.1702b98170614f@usc.edu>
Message-ID: <200304210454.h3L4sLxZ018897@dash.isi.edu>

On Sun, 20 Apr 2003 19:36:27 PDT, abhishek rajgarhia wrote: 
>Hi,
>
>Is the order in which the answer route queries are logged in s5.out, important? Does the order have to match the order of the queries in the input file?
>
>Thanks
>
>Abhishek 

In general, reordering input is not allowed, since it can change the
result.

Output is expected to follow from the input.

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 05:56:40 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 21:56:40 -0700
Subject: [Csci551-talk] A question about STL List of pointers ...
In-Reply-To: <001401c307b1$2a2a31c0$6401a8c0@darkstar>
Message-ID: <200304210456.h3L4uerl018946@dash.isi.edu>

On Sun, 20 Apr 2003 19:53:19 PDT, "Siddharth Shetye" wrote: 
>Hi,
>
>I was wondering ... if I have the following ...
>
>list<myclass*> list_of_my_class;
>
>it would be a list of pointer to type "myclass" ... so would I be required
>to code additional functions into "myclass" (as required by STL objects ,
>http://www.msoe.edu/eecs/ce/courseinfo/stl/objreq.htm) ??
>
>I was thinking that since they are a list of only pointers, c++/STL should
>be able to handle them natively (and maybe increment etc on the basis of
>sizeof(myclass) (or similar) )
>
>Thanks !
>Siddharth

Why don't you try it and see?

(Although I know STL compiler error messages are usually somewhere
between bad and awful, I think it will answer your question.)

   -John Heidemann

From xiw@usc.edu  Mon Apr 21 05:55:28 2003
From: xiw@usc.edu (Xi Wang)
Date: Sun, 20 Apr 2003 21:55:28 -0700
Subject: [Csci551-talk] standard answer
Message-ID: <028701c307c2$3c54b4e0$efe57d80@xwl3>

Hi. Can we get the latest version of standard output s7-1.out?

In project B document:
....
received packet id 4 group 225.0.101.15 from 3
rejected 224.0.11.1 due to unroutable RP
generating new packet id 4 for 224.0.10.1
received packet id 7 group 225.0.101.15 from 3
generating new packet id 8 for 225.0.101.15
sending packet id 8 group 225.0.101.15 to 3

Packet id 4 received before generated?
Looks like packet id 7 should be generated by 1, not received by 1 as
shown above.

Thanks,

Xi


From sjv@usc.edu  Mon Apr 21 06:19:40 2003
From: sjv@usc.edu (siddharth vora)
Date: Sun, 20 Apr 2003 22:19:40 -0700
Subject: [Csci551-talk] TTL in project B
Message-ID: <1787d8b1788850.17888501787d8b@usc.edu>

 Hi,

 My question is, If every node maintain the number of packets sent by a particular SENDER we can detect the loop.


 But, in this project we are not going to transfer from Shared tree to Source specific tree so, I think we have to implement TTL to prevent the loop.

Please correct me if I am wrong !!

Siddharth.
2632 Ellendle St.,
Appt. # 117
LA, CA 90007
Phone : 323 737 7522

----- Original Message -----
From: John Heidemann <johnh@ISI.EDU>
Date: Sunday, April 20, 2003 9:49 pm
Subject: Re: [Csci551-talk] TTL in project B

> On Sun, 20 Apr 2003 11:01:11 PDT, jigesh vora wrote: 
> >Hi,
> >  since we are using only muticast trees for communication there will not 
> be any loops. So can someone tell me the use of TTL in the messages?
> >
> >Thanx
> >Jigesh
> 
> You're correct that correct code will not have any loops.
> 
> All students in the class who can guarantee that all their code will
> work completely correctly the first time are not required to implement
> TTLs.
> 
> For the rest of the class, you might find that TTLs prevent
> catestrophic meltdowns if you accidently get a loop while you're
> testing your code.  (Think about it: what happens if you accidently
> get a loop where the packets get sent to multiple places... every time
> they loop back.)
> 
>   -John Heidemann
> 


From johnh@ISI.EDU  Mon Apr 21 06:36:17 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Sun, 20 Apr 2003 22:36:17 -0700
Subject: [Csci551-talk] homework 3 has been posted to the class web site
Message-ID: <200304210536.h3L5aHe3019594@dash.isi.edu>

Homework 3 has been posted to the class web site.
   -John Heidemann

From shshanks@aludra.usc.edu  Mon Apr 21 07:40:47 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Sun, 20 Apr 2003 23:40:47 -0700 (PDT)
Subject: [Csci551-talk] broken link ?
Message-ID: <Pine.GSO.4.33.0304202339460.23939-100000@aludra.usc.edu>

The newtests.tar.gz link on the TA website seems broken.
  -Sharma.


From gnawali@usc.edu  Mon Apr 21 07:50:37 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Sun, 20 Apr 2003 23:50:37 -0700
Subject: [Csci551-talk] broken link ?
In-Reply-To: Your message of Sun, 20 Apr 2003 23:40:47 -0700 (PDT)
Message-ID: <200304210650.h3L6obBb004372@enl.usc.edu>

Fixed.

you wrote:
> 
> The newtests.tar.gz link on the TA website seems broken.
>   -Sharma.

From nimeshmshah@hotmail.com  Mon Apr 21 08:21:39 2003
From: nimeshmshah@hotmail.com (nimesh shah-alias-nagda)
Date: Mon, 21 Apr 2003 00:21:39 -0700
Subject: [Csci551-talk] sample output files
Message-ID: <17f959417fbe42.17fbe4217f9594@usc.edu>

Hi Omprakash 

Can you please post sample input and output files for project B,
so that we can check with our output.

Thanks
Nimesh 


From ggrewal@usc.edu  Mon Apr 21 08:26:28 2003
From: ggrewal@usc.edu (gurpreet grewal)
Date: Mon, 21 Apr 2003 00:26:28 -0700
Subject: [Csci551-talk] Input file format
Message-ID: <17fb95b17fbec9.17fbec917fb95b@usc.edu>

Dear Professor,

I am assuming the following input file format(stage-6 and 7) for parsing:

1)Stage-6:
  After external routes I am given "route lookup" entries and after that I have 
  "mcast establishment" entries.

2)Stage-7:
  After external routes I am given "route lookup" entries and after that I have 
  "mcast establishment and sending" entries.

Please correct me if I am wrong.
  
Thanks
Gurpreet


From shshanks@aludra.usc.edu  Mon Apr 21 09:03:34 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Mon, 21 Apr 2003 01:03:34 -0700 (PDT)
Subject: [Csci551-talk] Input file format
In-Reply-To: <17fb95b17fbec9.17fbec917fb95b@usc.edu>
Message-ID: <Pine.GSO.4.33.0304210054130.8383-100000@aludra.usc.edu>

I would expect the following.
Atleast the delimiters for "all" stages are present in the file, whether
we run a particular stage or not.

Like:

After stage 4 input: we have,

0 0 0.0.0.0 <------------ for end of route lookups, stage 5
j 0 0.0.0.0    0.0.0.0  <---------- for end of joins, stage 6
j 0 0.0.0.0   <--------- for end of joins and sends, stage 7

Is this alright ?
I feel one syntactically-correct _complete_ input for
each of stages 5,6 and 7 would be helpful on the class webpages, so we
know what to expect and the grader doesnt have to worry about different
cfg-file syntax assumptions that people might make.
Thanks.

 -Sharma
___________________________________________

On Mon, 21 Apr 2003, gurpreet grewal wrote:

> Dear Professor,
>
> I am assuming the following input file format(stage-6 and 7) for parsing:
>
> 1)Stage-6:
>   After external routes I am given "route lookup" entries and after that I have
>   "mcast establishment" entries.
>
> 2)Stage-7:
>   After external routes I am given "route lookup" entries and after that I have
>   "mcast establishment and sending" entries.
>
> Please correct me if I am wrong.
>
> Thanks
> Gurpreet
>
>


From gpatel@usc.edu  Mon Apr 21 09:40:56 2003
From: gpatel@usc.edu (gaurav patel)
Date: Mon, 21 Apr 2003 01:40:56 -0700
Subject: [Csci551-talk] understanding stage 7 output requirements
Message-ID: <1824a791824a56.1824a561824a79@usc.edu>

Hi,
   The documentation for stage 7 output states "SID & PID are source and destination router ID's". 
   Is source the router that generated the data packet or the (last hop) router which forwarded the packet to receiving router ?
-Gaurav


From shetye@usc.edu  Mon Apr 21 11:30:46 2003
From: shetye@usc.edu (Siddharth Shetye)
Date: Mon, 21 Apr 2003 03:30:46 -0700
Subject: [Csci551-talk] multicast IP address range
References: <200304191905.h3JJ51028091@gamma.isi.edu>
Message-ID: <001301c307f1$1279f0c0$6401a8c0@darkstar>

Hi,

Isin't the range for valid multicast addresses from 224.0.0.0 to
239.255.255.255 ? If so then should we reject the request at router 3 which
has an address of 255.0.101.14 (which we are currently allowing) ? Or am I
wrong/missing something ?

Siddharth

> 1. In Section 3, change the statement
> "...or if the multicast address is not in the IP address range"
> to
> "...or if the multicast address is not in the IP multicast address range"
> [Add multicast]
>
> (You can assume that all addresses will be valid IP addresses.)


From shshanks@aludra.usc.edu  Mon Apr 21 20:33:27 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Mon, 21 Apr 2003 12:33:27 -0700 (PDT)
Subject: [Csci551-talk] any special format ?
Message-ID: <Pine.GSO.4.33.0304211231540.11140-100000@aludra.usc.edu>

The config file expects non-routable-messgs to be in the format:

rejected <a.b.c.d> due to unroutable RP

Any _required_ format for the multicast-address-out-of-range condition ?

  -Sharma


From johnh@ISI.EDU  Mon Apr 21 21:26:27 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 13:26:27 -0700
Subject: [Csci551-talk] updated project B
Message-ID: <200304212026.h3LKQRbj005526@dash.isi.edu>

Because of the number of updates to Project B, I've updated
the assignment on my web page with annotations for some of the
changes.

   -John Heidemann

From sbhavsar@usc.edu  Mon Apr 21 23:13:24 2003
From: sbhavsar@usc.edu (siddharth bhavsar)
Date: Mon, 21 Apr 2003 15:13:24 -0700
Subject: [Csci551-talk] Query about updated project B
Message-ID: <1b1bb491b1c4c3.1b1c4c31b1bb49@usc.edu>

Sir ,
the updated project specs says the 
s6-3.out should be :
joining 225.0.123.255 due to request from 5
joining 255.0.101.14 due to request from 3

But  should'nt  it also include the line 
joining 225.0.101.15 due to request from 3 ?

My reasoning is :
225.0.101.14 and 225.0.101.15 both have RP as the Router 4 
Though 3 send join to router 4 and also logs it in this file while creating an entry for the Multicast group 255.0.101.14 .
Now the request ;
j 3 192.168.101.15 225.0.101.15
implies that there is a new entry to be created for the multicast address 225.0.101.15 .There-fore it needs to be logged .and also 3 shoud send the join requsest message to 4 for the  new group 225.0.101.15 .
Correct me if i am wrong .

Also in the message
joining ABC due to request from XYZ
should XYZ be the immediate sender or the originator of that join request ?

Siddharth


----- Original Message -----
From: John Heidemann <johnh@ISI.EDU>
Date: Monday, April 21, 2003 1:26 pm
Subject: [Csci551-talk] updated project B

> 
> Because of the number of updates to Project B, I've updated
> the assignment on my web page with annotations for some of the
> changes.
> 
>   -John Heidemann
> 


From johnh@ISI.EDU  Mon Apr 21 17:52:04 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 09:52:04 -0700
Subject: [Csci551-talk] Input file format
In-Reply-To: <17fb95b17fbec9.17fbec917fb95b@usc.edu>
Message-ID: <200304211652.h3LGq4SD003959@dash.isi.edu>

On Mon, 21 Apr 2003 00:26:28 PDT, gurpreet grewal wrote: 
>Dear Professor,
>
>I am assuming the following input file format(stage-6 and 7) for parsing:
>
>1)Stage-6:
>  After external routes I am given "route lookup" entries and after that I have 
>  "mcast establishment" entries.
>
>2)Stage-7:
>  After external routes I am given "route lookup" entries and after that I have 
>  "mcast establishment and sending" entries.
>
>Please correct me if I am wrong.

That sounds consistent with the spec.

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 17:57:43 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 09:57:43 -0700
Subject: [Csci551-talk] multicast IP address range
In-Reply-To: <001301c307f1$1279f0c0$6401a8c0@darkstar>
Message-ID: <200304211658.h3LGvhLG003977@dash.isi.edu>

On Mon, 21 Apr 2003 03:30:46 PDT, "Siddharth Shetye" wrote: 
>Hi,
>
>Isin't the range for valid multicast addresses from 224.0.0.0 to
>239.255.255.255 ? If so then should we reject the request at router 3 which
>has an address of 255.0.101.14 (which we are currently allowing) ? Or am I
>wrong/missing something ?

That's a typo, it should be 225.0.101.14

   -John Heidemann

From johnh@ISI.EDU  Mon Apr 21 17:27:30 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 09:27:30 -0700
Subject: [Csci551-talk] TTL in project B
In-Reply-To: <1787d8b1788850.17888501787d8b@usc.edu>
Message-ID: <200304211628.h3LGRUxd003859@dash.isi.edu>

On Sun, 20 Apr 2003 22:19:40 PDT, siddharth vora wrote: 
>
> Hi,
>
> My question is, If every node maintain the number of packets sent by a particular SENDER we can detect the loop.
>
>
> But, in this project we are not going to transfer from Shared tree to Source specific tree so, I think we have to implement TTL to prevent the loop.
>
>Please correct me if I am wrong !!

This has only shared trees and doesn't require nodes to keep track of
packets set.

So I would encourage folks to implement TTLs as the specification says.

   -John Heidemann

>
>Siddharth.
>2632 Ellendle St.,
>Appt. # 117
>LA, CA 90007
>Phone : 323 737 7522
>
>----- Original Message -----
>From: John Heidemann <johnh@ISI.EDU>
>Date: Sunday, April 20, 2003 9:49 pm
>Subject: Re: [Csci551-talk] TTL in project B
>
>> On Sun, 20 Apr 2003 11:01:11 PDT, jigesh vora wrote: 
>> >Hi,
>> >  since we are using only muticast trees for communication there will not 
>> be any loops. So can someone tell me the use of TTL in the messages?
>> >
>> >Thanx
>> >Jigesh
>> 
>> You're correct that correct code will not have any loops.
>> 
>> All students in the class who can guarantee that all their code will
>> work completely correctly the first time are not required to implement
>> TTLs.
>> 
>> For the rest of the class, you might find that TTLs prevent
>> catestrophic meltdowns if you accidently get a loop while you're
>> testing your code.  (Think about it: what happens if you accidently
>> get a loop where the packets get sent to multiple places... every time
>> they loop back.)
>> 
>>   -John Heidemann
>> 


From johnh@ISI.EDU  Mon Apr 21 22:17:31 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 14:17:31 -0700
Subject: [Csci551-talk] any special format ?
In-Reply-To: <Pine.GSO.4.33.0304211231540.11140-100000@aludra.usc.edu>
Message-ID: <200304212117.h3LLHVtR005814@dash.isi.edu>

On Mon, 21 Apr 2003 12:33:27 PDT, Shshank Sharma wrote: 
>The config file expects non-routable-messgs to be in the format:
>
>rejected <a.b.c.d> due to unroutable RP
>
>Any _required_ format for the multicast-address-out-of-range condition ?
>

rejected <a.b.c.d> due to invalidate multicast group address


>  -Sharma


From amankapo@usc.edu  Mon Apr 21 23:51:09 2003
From: amankapo@usc.edu (aman kapoor)
Date: Mon, 21 Apr 2003 15:51:09 -0700
Subject: [Csci551-talk] Input file format
Message-ID: <1b5b34e1b570b8.1b570b81b5b34e@usc.edu>

Hi

In the project specifications it is mentioned in stage 6:

"Like stage 5, additional data will follow external routes, with each line beginning with an "j" followed by the router-id
where the join happens and the multicast-ip-address."

This means that in case of stage 6 there would not be

stage 5 delimiter 0 0 0.0.0.0 and stage 6 input would follow immediately after the external routes.

Thanks
Aman

----- Original Message -----
From: John Heidemann <johnh@ISI.EDU>
Date: Monday, April 21, 2003 9:52 am
Subject: Re: [Csci551-talk] Input file format

> On Mon, 21 Apr 2003 00:26:28 PDT, gurpreet grewal wrote: 
> >Dear Professor,
> >
> >I am assuming the following input file format(stage-6 and 7) for parsing:
> >
> >1)Stage-6:
> >  After external routes I am given "route lookup" entries and after that 
> I have 
> >  "mcast establishment" entries.
> >
> >2)Stage-7:
> >  After external routes I am given "route lookup" entries and after that 
> I have 
> >  "mcast establishment and sending" entries.
> >
> >Please correct me if I am wrong.
> 
> That sounds consistent with the spec.
> 
>   -John Heidemann
> 


From johnh@ISI.EDU  Mon Apr 21 23:55:58 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 15:55:58 -0700
Subject: [Csci551-talk] Input file format
In-Reply-To: <1b5b34e1b570b8.1b570b81b5b34e@usc.edu>
Message-ID: <200304212255.h3LMtwtT003160@dash.isi.edu>

On Mon, 21 Apr 2003 15:51:09 PDT, aman kapoor wrote: 
>Hi
>
>In the project specifications it is mentioned in stage 6:
>
>"Like stage 5, additional data will follow external routes, with each line beginning with an "j" followed by the router-id
>where the join happens and the multicast-ip-address."
>
>This means that in case of stage 6 there would not be
>
>stage 5 delimiter 0 0 0.0.0.0 and stage 6 input would follow immediately after the external routes.

Correct.  But the stage 6 input will end with a delimiter as well
(just like stage 5 does).

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 01:16:38 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 17:16:38 -0700
Subject: [Csci551-talk] Input file format
In-Reply-To: <Pine.GSO.4.33.0304210054130.8383-100000@aludra.usc.edu>
Message-ID: <200304220016.h3M0GcWs003645@dash.isi.edu>

On Mon, 21 Apr 2003 01:03:34 PDT, Shshank Sharma wrote: 
>
>I would expect the following.
>Atleast the delimiters for "all" stages are present in the file, whether
>we run a particular stage or not.
>
>Like:
>
>After stage 4 input: we have,
>
>0 0 0.0.0.0 <------------ for end of route lookups, stage 5
>j 0 0.0.0.0    0.0.0.0  <---------- for end of joins, stage 6
>j 0 0.0.0.0   <--------- for end of joins and sends, stage 7
>
>Is this alright ?
>I feel one syntactically-correct _complete_ input for
>each of stages 5,6 and 7 would be helpful on the class webpages, so we
>know what to expect and the grader doesnt have to worry about different
>cfg-file syntax assumptions that people might make.
>Thanks.

That's not what will happen.

The intent in the input file is that all of stage5/6/7 sections would
be compatible.

Any input would have stage4 input + other stuff,
where the other stuff is EITHER stage5/6/7.
(And it would be easy for your parser to handle any of the
alternatives.)
Thus your expectation that there be a stage5 terminator
AND a stage6 terminator AND a stage7 terminator is not consistent with
this approach.

Unfortunately, the current spec calls for two different terminators,
one for stage5 ("0 0 0.0.0.0") and the other for stage6/7 ("j 0
0.0.0.0 0.0.0.0").

I would suggest that it is probably easiest if you write your parser to
handle stage5/6/7 all in one place, and accept either termintor as
ending the file.  This approach should not cause problems for our
input files.

(But do not expect 3 separate terminators!)


We WILL post complete sample inputs to the web by tomorrow.

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 01:18:25 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 17:18:25 -0700
Subject: [Csci551-talk] Re: understanding stage 7 output requirements
In-Reply-To: <1824a791824a56.1824a561824a79@usc.edu>
Message-ID: <200304220018.h3M0IPi4003691@dash.isi.edu>

On Mon, 21 Apr 2003 01:40:56 PDT, gaurav patel wrote: 
>Hi,
>   The documentation for stage 7 output states "SID & PID are source and destination router ID's". 
>   Is source the router that generated the data packet or the (last hop) router which forwarded the packet to receiving router ?
>-Gaurav

Last hop router.

   -John Heidemann

From shshanks@aludra.usc.edu  Tue Apr 22 01:23:02 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Mon, 21 Apr 2003 17:23:02 -0700 (PDT)
Subject: [Csci551-talk] Input file format
In-Reply-To: <200304220016.h3M0GcWs003645@dash.isi.edu>
Message-ID: <Pine.GSO.4.33.0304211719140.11709-100000@aludra.usc.edu>

> Thus your expectation that there be a stage5 terminator
> AND a stage6 terminator AND a stage7 terminator is not consistent with
> this approach.

Yes, thanks. I read that in the updated spec.

> I would suggest that it is probably easiest if you write your parser to
> handle stage5/6/7 all in one place, and accept either termintor as
> ending the file.  This approach should not cause problems for our
> input files.

Did just that. Also noted that the delimiter for both stage 6 and 7 has
now been changed to
j 0 0.0.0.0 0.0.0.0

while it was j 0 0.0.0.0 earlier for stage 7.

Thanks.
 -Sharma



From johnh@ISI.EDU  Tue Apr 22 01:24:57 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Mon, 21 Apr 2003 17:24:57 -0700
Subject: [Csci551-talk] how to send data on the mcast tree
Message-ID: <200304220024.h3M0OvQG003742@dash.isi.edu>

I've seen some confusion about how to send data on the mcast tree in
Proj B.

A student asks:
>I have a question regarding sending packets in stage 7.
>
>In a topology like this: 1-2-3-4-5.
>
>If 3 is a RP for Group G, and 1,5 explicitly send the Join G message.
>(all nodes are on the tree now)
>
>Later on, node 2 wants to send out a data packet to G,  does 2 send it
>towards RP 3 and let 3 send out the data packet along the tree, Or 2 can
>send the data to node 1 directly?

1. In PIM-SM, if a node on the tree sends data, it just sends it to the
tree, not to the RP first.

2. In PIM-SM, if a node is NOT on the tree, it tunnels the packet to the
RP that sends it on the tree.

The spec says were are NOT implementing 2.  Therefore nodes not on the
tree that try to send data just drop the data.

Nodes that are on the tree should behave like 1.

   -John Heidemann

From gnawali@usc.edu  Tue Apr 22 07:13:56 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Mon, 21 Apr 2003 23:13:56 -0700
Subject: [Csci551-talk] HW2 notification
Message-ID: <200304220613.h3M6DuYQ016882@enl.usc.edu>

I just sent an email to everyone who submitted HW2. If you did not get
an email from me, you did not submit your HW2 or at least I did not
receive it and you should contact me immediately.

From rajgarhi@usc.edu  Tue Apr 22 08:17:27 2003
From: rajgarhi@usc.edu (abhishek rajgarhia)
Date: Tue, 22 Apr 2003 00:17:27 -0700
Subject: [Csci551-talk] Origin router
Message-ID: <203c79206a1d.206a1d203c79@usc.edu>

Hi,

Wanted to know what the definition of origin router is, in case of an external route. Is it the internal router Id which first injected the route, or the external router id?

If we look at the dv-message format, in the place for origin routers we do have external router ids. But the outputs obtained by running the code provided on the class web-page with stage4m.in suggest that external router id cannot be the origin router.

This is because for router 5 we have to paths to 192.168.16.0/16

(5 2 21 ) and (5 4 20) with same MED value and (5 2 21) is selected as the final route. This would be consistent if we considered the origin routers for the paths to be 2 and 4 respectively rather than 21 and 20.

Thank you

Regards

Abhishek



From sjv@usc.edu  Tue Apr 22 12:16:21 2003
From: sjv@usc.edu (siddharth vora)
Date: Tue, 22 Apr 2003 04:16:21 -0700
Subject: [Csci551-talk] Question regarding the whole project implementation !!!!
Message-ID: <262c2f2663e2.2663e2262c2f@usc.edu>

Hi,

 I am having a big trouble over here. Do we require to print output same as the GIVEN ??

 Because as per previous mail and per Spec we need to process input file sequence.

 I follow the sequence of the input file, but after that I dont maintain the MANAGER to wait for each request !!


 I sent out the request from manager to the appropriate router and then make router wait for the messages for which it doesnot get acknowledgements and once it gets the acknowledgement of its JOIN messages I send SEND messages to other routers !!

 But by doing this I am not able to maintain the OUTPUT sequence !!

Do, we need to maintain the OUTPUT sequence ?? Its urgent because I have to change my whole code if its not true. 

 Do we have to maintain the output sequence (Not the INPUT sequence) ?

 Please reply me as soon as possible.

Thanks,
Siddharth.

 
 
 Thanks,



2632 Ellendle St.,
Appt. # 117
LA, CA 90007
Phone : 323 737 7522




From shetye@usc.edu  Tue Apr 22 12:41:22 2003
From: shetye@usc.edu (Siddharth Shetye)
Date: Tue, 22 Apr 2003 04:41:22 -0700
Subject: [Csci551-talk] external nodes ...
References: <200304211905.h3LJ53015160@gamma.isi.edu>
Message-ID: <002201c308c4$197e5db0$6401a8c0@darkstar>

Currently what I'm doing is that if someone is trying to "join" a multicast
group with the RP as an external node, I am allowing those joins and am
using internal node connected to that external node to handle all the
multicast messages for the external one.

eg
                / B
              /
Ext-----A --- C
              \
                \ D

Internal Links: A-B , A-C, A-D
External Links: Ext-A
If RP = Ext then B,C,D all send traffic to A (the prefered route to Ext)
which in theory could forward traffic to ext.
This makes the ext RP "reachable" to join exit messages targeted at Ext.

My question is ... is this effort to make the external RP reachable is ok ?
or do we simply assume external nodes don't exist in stage 6 and 7 ?

Thanks
Siddharth
ZZZzzzzzzzzzzz...


From athavale@usc.edu  Tue Apr 22 12:49:12 2003
From: athavale@usc.edu (amar athavale)
Date: Tue, 22 Apr 2003 04:49:12 -0700
Subject: [Csci551-talk] Errors in the new ProjB handout
Message-ID: <26845326740a.26740a268453@usc.edu>

Hi,

   In the new Proj B doc, the sample s7-1 output has a couple of errors. 
1) On output line 9, instead of packet ID 4 it should be packet id 5. 
2) packet id 7 should be generated and sent, instead of receiving it from 2 (line 10).

Amar.


From sbhavsar@usc.edu  Tue Apr 22 13:47:26 2003
From: sbhavsar@usc.edu (siddharth bhavsar)
Date: Tue, 22 Apr 2003 05:47:26 -0700
Subject: [Csci551-talk] Errors in the new ProjB handout
Message-ID: <26f565271143.27114326f565@usc.edu>

Yes ,I second that .
I am assuming that the data packets  will be assigned sequentail PacketIds irrespetive of the router to which the manager will initially send it.That is , the manager will maintain a single counter for all s packets it dispatches rather than a counter for each  router to which it dispatches the s packets .
Is my assumtion for packet ids ok ?
Thanks 
Siddharth 

----- Original Message -----
From: amar athavale <athavale@usc.edu>
Date: Tuesday, April 22, 2003 4:49 am
Subject: [Csci551-talk] Errors in the new ProjB handout

> Hi,
> 
>   In the new Proj B doc, the sample s7-1 output has a couple of errors. 
> 1) On output line 9, instead of packet ID 4 it should be packet id 5. 
> 2) packet id 7 should be generated and sent, instead of receiving it from 
> 2 (line 10).
> 
> Amar.
> 
> 


From johnh@ISI.EDU  Tue Apr 22 19:42:53 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 11:42:53 -0700
Subject: [Csci551-talk] Origin router
In-Reply-To: <203c79206a1d.206a1d203c79@usc.edu>
Message-ID: <200304221842.h3MIgrfB005236@dash.isi.edu>

On Tue, 22 Apr 2003 00:17:27 PDT, abhishek rajgarhia wrote: 
>Hi,
>
>Wanted to know what the definition of origin router is, in case of an external route. Is it the internal router Id which first injected the route, or the external router id?
>
>If we look at the dv-message format, in the place for origin routers we do have external router ids. But the outputs obtained by running the code provided on the class web-page with stage4m.in suggest that external router id cannot be the origin router.
>
>This is because for router 5 we have to paths to 192.168.16.0/16
>
>(5 2 21 ) and (5 4 20) with same MED value and (5 2 21) is selected as the final route. This would be consistent if we considered the origin routers for the paths to be 2 and 4 respectively rather than 21 and 20.

This was not completely specified in Project A,
so you can make either choice, provided you justified it in your
README.

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 19:48:22 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 11:48:22 -0700
Subject: [Csci551-talk] Question regarding the whole project implementation !!!!
In-Reply-To: <262c2f2663e2.2663e2262c2f@usc.edu>
Message-ID: <200304221848.h3MImMR7005291@dash.isi.edu>

On Tue, 22 Apr 2003 04:16:21 PDT, siddharth vora wrote: 
>Hi,
>
> I am having a big trouble over here. Do we require to print output same as the GIVEN ??
>
> Because as per previous mail and per Spec we need to process input file sequence.
>
> I follow the sequence of the input file, but after that I dont maintain the MANAGER to wait for each request !!
>
>
> I sent out the request from manager to the appropriate router and then make router wait for the messages for which it doesnot get acknowledgements and once it gets the acknowledgement of its JOIN messages I send SEND messages to other routers !!
>
> But by doing this I am not able to maintain the OUTPUT sequence !!
>
>Do, we need to maintain the OUTPUT sequence ?? Its urgent because I have to change my whole code if its not true. 
>
> Do we have to maintain the output sequence (Not the INPUT sequence) ?

It's not obvious to me why you cannot have the output sequence match
the input sequence, but the spec says

	You should also allow sufficient time to ensure a join occurs between
	each step of the input; for most implementations a few seconds should
	be plenty.

The purpose of this is that the output sequence matches the input
sequence.

If you do not do this, you should document that fact and why you do
not in the README.  (And you would not be meeting the spec.)


   -John Heidemann


>
> Please reply me as soon as possible.
>
>Thanks,
>Siddharth.
>
> 
> 
> Thanks,
>
>
>
>2632 Ellendle St.,
>Appt. # 117
>LA, CA 90007
>Phone : 323 737 7522
>
>


From johnh@ISI.EDU  Tue Apr 22 19:57:26 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 11:57:26 -0700
Subject: [Csci551-talk] external nodes ...
In-Reply-To: <002201c308c4$197e5db0$6401a8c0@darkstar>
Message-ID: <200304221857.h3MIvQeN005371@dash.isi.edu>

On Tue, 22 Apr 2003 04:41:22 PDT, "Siddharth Shetye" wrote: 
>Currently what I'm doing is that if someone is trying to "join" a multicast
>group with the RP as an external node, I am allowing those joins and am
>using internal node connected to that external node to handle all the
>multicast messages for the external one.
>
>eg
>                / B
>              /
>Ext-----A --- C
>              \
>                \ D
>
>Internal Links: A-B , A-C, A-D
>External Links: Ext-A
>If RP = Ext then B,C,D all send traffic to A (the prefered route to Ext)
>which in theory could forward traffic to ext.
>This makes the ext RP "reachable" to join exit messages targeted at Ext.
>
>My question is ... is this effort to make the external RP reachable is ok ?
>or do we simply assume external nodes don't exist in stage 6 and 7 ?
>
>Thanks
>Siddharth
>ZZZzzzzzzzzzzz...

As posted yesterday, you do not need to support RPs on external nodes.

There will be no RP addresses routed by external routers.
They will all be either internal routers or unrouted address space.

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 21:51:14 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 13:51:14 -0700
Subject: [Csci551-talk] Errors in the new ProjB handout
In-Reply-To: <26845326740a.26740a268453@usc.edu>
Message-ID: <200304222051.h3MKpEKO006123@dash.isi.edu>

On Tue, 22 Apr 2003 04:49:12 PDT, amar athavale wrote: 
>Hi,
>
>   In the new Proj B doc, the sample s7-1 output has a couple of errors. 
>1) On output line 9, instead of packet ID 4 it should be packet id 5. 
>2) packet id 7 should be generated and sent, instead of receiving it from 2 (line 10).
>
>Amar.


You are correct; these are errors in the handout.

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 21:51:50 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 13:51:50 -0700
Subject: [Csci551-talk] Errors in the new ProjB handout
In-Reply-To: <26f565271143.27114326f565@usc.edu>
Message-ID: <200304222051.h3MKpoHt006144@dash.isi.edu>

On Tue, 22 Apr 2003 05:47:26 PDT, siddharth bhavsar wrote: 
>Yes ,I second that .

>I am assuming that the data packets  will be assigned sequentail PacketIds irrespetive of the router to which the manager will initially send it.That is , the manager will maintain a single counter for all s packets it dispatches rather than a counter for each  router to which it dispatches the s packets .
>Is my assumtion for packet ids ok ?

That is a reasonable way to generate packet ids.

   -John Heidemann

From johnh@ISI.EDU  Tue Apr 22 21:57:45 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 13:57:45 -0700
Subject: [Csci551-talk] another update to project b spec
Message-ID: <200304222057.h3MKvjoN006257@dash.isi.edu>

Another two small bug fixes to sample outputs have been posted to the
Project B specification.

In addition, Amol has generated a more complete sample output for
Project B.  THat is available for download from my web page.
This output was generated by hand and so it's possible it has errors,
but you may find it useful.

   -John Heidemann


From bindlish@usc.edu  Wed Apr 23 00:47:31 2003
From: bindlish@usc.edu (aarti bindlish)
Date: Tue, 22 Apr 2003 16:47:31 -0700
Subject: [Csci551-talk] Doubts in stage 6 & 7 output
Message-ID: <5d1d865cebca.5cebca5d1d86@usc.edu>

Hi!
I have two doubts:
1. In stage 6 output(as given in handout), for s6-3.out, I think it should reject the join request to 255.0.101.14 because this address is out of multicast IP address range.(which is from 224.0.0.0 to 239.255.255.255). 

2. In stage 7 output, for s7-1.out, acc. to handout the 8th and 9th lines are:

rejected 224.0.11.1 due to unroutable RP
generating new packet id 5 for 224.0.10.1

But I am getting these two lines in reverse order. I.e.

generating new packet id 5 for 224.0.10.1
rejected 224.0.11.1 due to unroutable RP

Did anyone get the same output? Pls let me know if I am wrong.


Thanks,
Aarti Bindlish
MS(Computer Science)
University of Southern California
Los Angeles
Ph: 323-733-4010


From johnh@ISI.EDU  Wed Apr 23 01:53:47 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 17:53:47 -0700
Subject: [Csci551-talk] Query about updated project B
In-Reply-To: <1b1bb491b1c4c3.1b1c4c31b1bb49@usc.edu>
Message-ID: <200304230053.h3N0rmon007791@dash.isi.edu>

On Mon, 21 Apr 2003 15:13:24 PDT, siddharth bhavsar wrote: 
>Sir ,
>the updated project specs says the 
>s6-3.out should be :
>joining 225.0.123.255 due to request from 5
>joining 255.0.101.14 due to request from 3
>
>But  should'nt  it also include the line 
>joining 225.0.101.15 due to request from 3 ?

Yes.

>My reasoning is :
>225.0.101.14 and 225.0.101.15 both have RP as the Router 4 
>Though 3 send join to router 4 and also logs it in this file while creating an entry for the Multicast group 255.0.101.14 .
>Now the request ;
>j 3 192.168.101.15 225.0.101.15
>implies that there is a new entry to be created for the multicast address 225.0.101.15 .There-fore it needs to be logged .and also 3 shoud send the join requsest message to 4 for the  new group 225.0.101.15 .
>Correct me if i am wrong .
>
>Also in the message
>joining ABC due to request from XYZ
>should XYZ be the immediate sender or the originator of that join request ?

Immediate sender.

   -John Heidemann

>
>Siddharth
>
>
>----- Original Message -----
>From: John Heidemann <johnh@ISI.EDU>
>Date: Monday, April 21, 2003 1:26 pm
>Subject: [Csci551-talk] updated project B
>
>> 
>> Because of the number of updates to Project B, I've updated
>> the assignment on my web page with annotations for some of the
>> changes.
>> 
>>   -John Heidemann
>> 


From gpatel@usc.edu  Wed Apr 23 03:59:16 2003
From: gpatel@usc.edu (gaurav patel)
Date: Tue, 22 Apr 2003 19:59:16 -0700
Subject: [Csci551-talk] submission of the code
Message-ID: <6ae9d56afa1e.6afa1e6ae9d5@usc.edu>

Hi,
   Those who have used the professor's code, how is the "timers" directory to be submitted?

 When submitting like 

submit -user csci551 -tag projb manager.cc ... /timers/timers.cc ....

will, during submission, a timers directory be created?
How is the submission to be carried out?
-Gaurav


From xiw@usc.edu  Wed Apr 23 04:09:40 2003
From: xiw@usc.edu (Xi Wang)
Date: Tue, 22 Apr 2003 20:09:40 -0700
Subject: [Csci551-talk] submission of the code
References: <6ae9d56afa1e.6afa1e6ae9d5@usc.edu>
Message-ID: <00eb01c30945$cc7cf300$41e47d80@xwl3>

I put all files in a single directory. Just change your makefile. This
would avoid potential problems.

Xi

----- Original Message -----
From: "gaurav patel" <gpatel@usc.edu>
To: "csci551 mailgroup -" <csci551-talk@mailman.isi.edu>
Cc: <johnh@ISI.EDU>
Sent: Tuesday, April 22, 2003 7:59 PM
Subject: [Csci551-talk] submission of the code


> Hi,
>    Those who have used the professor's code, how is the "timers"
directory to be submitted?
>
>  When submitting like
>
> submit -user csci551 -tag projb manager.cc ... /timers/timers.cc ....
>
> will, during submission, a timers directory be created?
> How is the submission to be carried out?
> -Gaurav
>
>


From shshanks@aludra.usc.edu  Wed Apr 23 04:22:33 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Tue, 22 Apr 2003 20:22:33 -0700 (PDT)
Subject: [Csci551-talk] missing lines in s6-4.out (sample o/p)
Message-ID: <Pine.GSO.4.33.0304222019470.26920-100000@aludra.usc.edu>

s6-4.out from the sample outputs on the webpage is apparently missing
two lines:

Joining 225.0.101.15 due to request from 4
Joining 225.0.101.15 due to request from 3
 - Sharma
________________________________________________________


From shshanks@aludra.usc.edu  Wed Apr 23 04:56:52 2003
From: shshanks@aludra.usc.edu (Shshank Sharma)
Date: Tue, 22 Apr 2003 20:56:52 -0700 (PDT)
Subject: [Csci551-talk] missing lines in s6-4.out (sample o/p)
In-Reply-To: <Pine.GSO.4.33.0304222019470.26920-100000@aludra.usc.edu>
Message-ID: <Pine.GSO.4.33.0304222049140.7272-100000@aludra.usc.edu>

On Tue, 22 Apr 2003, Shshank Sharma wrote:
>
> s6-4.out from the sample outputs on the webpage is apparently missing
> two lines:
>
> Joining 225.0.101.15 due to request from 4
> Joining 225.0.101.15 due to request from 3


Although I feel logging _all_ join requests for a particular mcast group
(irrespective of whether the logging-node is on the tree already or not)
is more descriptive, the Professor's comment at:
http://mailman.isi.edu/pipermail/csci551-talk/2003-April/000516.html
does not necessitate that.

Are both ok ?

 -Sharma
___________________________________________



From gnawali@usc.edu  Wed Apr 23 05:25:59 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Tue, 22 Apr 2003 21:25:59 -0700
Subject: [Csci551-talk] slip day request procedure
Message-ID: <200304230425.h3N4Pxbi028783@enl.usc.edu>

Send an email to gnawali@usc.edu with the following subject line:

CSCI551 ProjB Slip Day Request



You do not need to put any text in the body. And I will ignore all the
requests that does not follow this guideline.

From gnawali@usc.edu  Wed Apr 23 05:28:40 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Tue, 22 Apr 2003 21:28:40 -0700
Subject: [Csci551-talk] submission of the code
In-Reply-To: Your message of Tue, 22 Apr 2003 20:09:40 -0700
Message-ID: <200304230428.h3N4Se0U028842@enl.usc.edu>

Please collect all the files you use in one directory (including the
timers file you use) and submit them all. A good way to test if your
submission is self contained is by copying the files you are
submitting to a different directory and making it there.

Please do not submit the binaries and other files. If you do, it shows
that you are not thinking about what you are submitting.

Good luck.

you wrote:
> I put all files in a single directory. Just change your makefile. This
> would avoid potential problems.
> 
> Xi
> 
> ----- Original Message -----
> From: "gaurav patel" <gpatel@usc.edu>
> To: "csci551 mailgroup -" <csci551-talk@mailman.isi.edu>
> Cc: <johnh@ISI.EDU>
> Sent: Tuesday, April 22, 2003 7:59 PM
> Subject: [Csci551-talk] submission of the code
> 
> 
> > Hi,
> >    Those who have used the professor's code, how is the "timers"
> directory to be submitted?
> >
> >  When submitting like
> >
> > submit -user csci551 -tag projb manager.cc ... /timers/timers.cc ....
> >
> > will, during submission, a timers directory be created?
> > How is the submission to be carried out?
> > -Gaurav
> >
> >

From johnh@ISI.EDU  Wed Apr 23 05:56:58 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 21:56:58 -0700
Subject: [Csci551-talk] submission of the code
In-Reply-To: <6ae9d56afa1e.6afa1e6ae9d5@usc.edu>
Message-ID: <200304230456.h3N4uxl5002909@dash.isi.edu>

On Tue, 22 Apr 2003 19:59:16 PDT, gaurav patel wrote: 
>Hi,
>   Those who have used the professor's code, how is the "timers" directory to be submitted?
>
> When submitting like 
>
>submit -user csci551 -tag projb manager.cc ... /timers/timers.cc ....
>
>will, during submission, a timers directory be created?
>How is the submission to be carried out?
>-Gaurav

This is covered in section 6 of the proj b handout.

   -John Heidemann

From johnh@ISI.EDU  Wed Apr 23 06:03:30 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Tue, 22 Apr 2003 22:03:30 -0700
Subject: [Csci551-talk] Re: query regarding s7-output
In-Reply-To: <002c01c3094d$ba305f90$d8f97d80@Sumeet>
Message-ID: <200304230503.h3N53U1k003095@dash.isi.edu>

On Tue, 22 Apr 2003 21:06:33 PDT, Sumeet Savla wrote: 
>[1  <text/plain; iso-8859-1 (7bit)>]
>
>[2  <text/html; iso-8859-1 (7bit)>]
>Hi professor,
> 
>Just needed a confirmation on what i am doing...
>whenever i send a packet, i am logging it into the file.
>So, this means that if I am resending a packet, i log it again as "sending
>packet....".
> 
>Same happens on the receive side....
>i.e. if I receive the same packet again (due to an ack being dropped), i log
>it again as "receiving packet....".
> 
> 
>Sumeet
>
>

That sounds OK, although I'd document your choices there in the
README.

   -John Heidemann

From gpatel@usc.edu  Wed Apr 23 09:00:39 2003
From: gpatel@usc.edu (gaurav patel)
Date: Wed, 23 Apr 2003 01:00:39 -0700
Subject: [Csci551-talk] Re: query regarding s7-output
Message-ID: <7e1bfc7e29bb.7e29bb7e1bfc@usc.edu>


----- Original Message -----
From: John Heidemann <johnh@ISI.EDU>
Date: Tuesday, April 22, 2003 10:03 pm
Subject: [Csci551-talk] Re: query regarding s7-output

> On Tue, 22 Apr 2003 21:06:33 PDT, Sumeet Savla wrote: 
> >[1  <text/plain; iso-8859-1 (7bit)>]
> >
> >[2  <text/html; iso-8859-1 (7bit)>]
> >Hi professor,
> > 
> >Just needed a confirmation on what i am doing...
> >whenever i send a packet, i am logging it into the file.
> >So, this means that if I am resending a packet, i log it again as "sending
> >packet....".
> > 
> >Same happens on the receive side....
> >i.e. if I receive the same packet again (due to an ack being dropped), i log
> >it again as "receiving packet....".
> > 
> > 
> >Sumeet
> >
> >
> 
> That sounds OK, although I'd document your choices there in the
> README.

Irrespective of choice, as per the requirements, do we HAVE to log the retansmission & duplicate messages received?


>   -John Heidemann
> 


From sumeetsavla@lycos.co.uk  Wed Apr 23 13:10:04 2003
From: sumeetsavla@lycos.co.uk (Sumeet Savla)
Date: Wed, 23 Apr 2003 13:10:04 +0100
Subject: [Csci551-talk] Dropping acks?
Message-ID: <1051096204005290@lycos-europe.com>

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_Lycos_0052901051096204_ID
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

<STYLE>TABLE.mail_content {DISPLAY: block}</STYLE><TABLE width="440" height="300" Class="mail_content" ALIGN="CENTER"><TR><TD VALIGN="TOP" style="position:relative;"><DIV><FONT face=Arial size=2>Hi</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>I have made all udp traffic reliable (both the joins and the sends).</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>I have a doubt regarding what should happen if an acknowledgement packet is dropped.</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>The timeout will occur due to lack of acknowledgement and the packet will be retransmitted.</FONT></DIV>
<DIV><FONT face=Arial size=2>However, since we had already sent the packet which had been "received" by the routers (and processed too);</FONT></DIV>
<DIV><FONT face=Arial size=2>what do we do about this resent packet - is it okay if we reprocess it or should it be ignored?</FONT></DIV>
<DIV><FONT face=Arial size=2>Are both methods okay??</FONT></DIV>
<DIV><FONT face=Arial size=2>I am not sure if the reprocessing leads to looping , because it ddint go into loops in my test-cases!</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>Or should I simply not drop acks and keep it simple?</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>Sumeet</FONT></DIV></TD></TR></TABLE><P><a href="http://mail.lycos.co.uk">Get your free 15MB mailbox on Lycos Email</a>


--=_NextPart_Lycos_0052901051096204_ID--


From sumeetsavla@lycos.co.uk  Wed Apr 23 13:11:18 2003
From: sumeetsavla@lycos.co.uk (Sumeet Savla)
Date: Wed, 23 Apr 2003 13:11:18 +0100
Subject: [Csci551-talk] (no subject)
Message-ID: <1051096278006141@lycos-europe.com>

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_Lycos_0061411051096278_ID
Content-Type: text/html; charset="iso-8859-1"
Content-Transfer-Encoding: 7bit

<STYLE>TABLE.mail_content {DISPLAY: block}</STYLE><TABLE width="440" height="300" Class="mail_content" ALIGN="CENTER"><TR><TD VALIGN="TOP" style="position:relative;">Hi professor...<BR>This email defnitely increases the confusion!!!<BR>PLease DO clarify if these 2 lines are missing or these should not be<BR>logged...<BR><BR>Sumeet<BR>----- Original Message -----<BR>From: "Shshank Sharma" &lt;<A href="mailto:shshanks@aludra.usc.edu">shshanks@aludra.usc.edu</A>&gt;<BR>To: &lt;<A href="mailto:csci551-talk@mailman.isi.edu">csci551-talk@mailman.isi.edu</A>&gt;<BR>Sent: Tuesday, April 22, 2003 8:56 PM<BR>Subject: Re: [Csci551-talk] missing lines in s6-4.out (sample o/p)<BR><BR><BR><BR>On Tue, 22 Apr 2003, Shshank Sharma wrote:<BR>&gt;<BR>&gt; s6-4.out from the sample outputs on the webpage is apparently missing<BR>&gt; two lines:<BR>&gt;<BR>&gt; Joining 225.0.101.15 due to request from 4<BR>&gt; Joining 225.0.101.15 due to request from 3<BR><BR><BR><BR></TD></TR></TABLE><P><a href=
 "http://mail.lycos.co.uk">Get your free 15MB mailbox on Lycos Email</a>


--=_NextPart_Lycos_0061411051096278_ID--


From athavale@usc.edu  Wed Apr 23 12:32:17 2003
From: athavale@usc.edu (amar athavale)
Date: Wed, 23 Apr 2003 04:32:17 -0700
Subject: [Csci551-talk] (no subject)
Message-ID: <81dd9c81dafa.81dafa81dd9c@usc.edu>

This is a multi-part message in MIME format.

--Boundary_(ID_358ldTR1ESmTORIMEYJcOQ)
Content-type: text/plain; charset=us-ascii
Content-transfer-encoding: 7BIT
Content-disposition: inline

Hi,

   This question has been answered in one of the previous mails. If a router is a member of a group, then there is no need to log it. Why would, say an RP join the same group n times, because it receives a join request from (n neighbours + itself). 

Amar.



--Boundary_(ID_358ldTR1ESmTORIMEYJcOQ)
Content-type: text/html; charset=iso-8859-1
Content-transfer-encoding: 7BIT

<STYLE>TABLE.mail_content {DISPLAY: block}</STYLE><TABLE width="440" height="300" Class="mail_content" ALIGN="CENTER"><TR><TD VALIGN="TOP" style="position:relative;">Hi professor...<BR>This email defnitely increases the confusion!!!<BR>PLease DO clarify if these 2 lines are missing or these should not be<BR>logged...<BR><BR>Sumeet<BR>----- Original Message -----<BR>From: "Shshank Sharma" &lt;<A href="mailto:shshanks@aludra.usc.edu">shshanks@aludra.usc.edu</A>&gt;<BR>To: &lt;<A href="mailto:csci551-talk@mailman.isi.edu">csci551-talk@mailman.isi.edu</A>&gt;<BR>Sent: Tuesday, April 22, 2003 8:56 PM<BR>Subject: Re: [Csci551-talk] missing lines in s6-4.out (sample o/p)<BR><BR><BR><BR>On Tue, 22 Apr 2003, Shshank Sharma wrote:<BR>&gt;<BR>&gt; s6-4.out from the sample outputs on the webpage is apparently missing<BR>&gt; two lines:<BR>&gt;<BR>&gt; Joining 225.0.101.15 due to request from 4<BR>&gt; Joining 225.0.101.15 due to request from 3<BR><BR><BR><BR></TD></TR></TABLE><P><a href=
 "http://mail.lycos.co.uk">Get your free 15MB mailbox on Lycos Email</a>


--Boundary_(ID_358ldTR1ESmTORIMEYJcOQ)--

From gnawali@usc.edu  Wed Apr 23 19:45:45 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 23 Apr 2003 11:45:45 -0700
Subject: [Csci551-talk] after project submission
Message-ID: <200304231845.h3NIjjQu016913@enl.usc.edu>

Dear class,

After the project submission deadline (after noon today), I will:

* Send out an email to students who submitted the project by the
deadline. If you don't receive this email by lets say 1pm, you should
contact me immediately.

* Send out an email to students whose code I can not "make" in
aludra. If you get this email, it is your responsibility to make sure
your code compiles in aludra and resubmit. Depending on when you
resubmit your project, we will count this as a project late by the
number of days it takes you to fix the code.

* Send out an email to students who requested slip days. It is very
inefficient for me to process slip days one at a time. I will
batch-process them all and send emails.

Good luck.

- om_p

From johnh@ISI.EDU  Wed Apr 23 21:45:34 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 23 Apr 2003 13:45:34 -0700
Subject: [Csci551-talk] Doubts in stage 6 & 7 output
In-Reply-To: <5d1d865cebca.5cebca5d1d86@usc.edu>
Message-ID: <200304232045.h3NKjYak021452@dash.isi.edu>

On Tue, 22 Apr 2003 16:47:31 PDT, aarti bindlish wrote: 
>Hi!
>I have two doubts:
>1. In stage 6 output(as given in handout), for s6-3.out, I think it should reject the join request to 255.0.101.14 because this address is out of multicast IP address range.(which is from 224.0.0.0 to 239.255.255.255). 

I had previously posted that the 255.0.101.14 was a typo and fixed it
in an update to project b.

But you're right, with the address, it shoudl be rejected.


From gnawali@usc.edu  Wed Apr 23 21:49:41 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Wed, 23 Apr 2003 13:49:41 -0700
Subject: [Csci551-talk] Project B submission status
Message-ID: <200304232049.h3NKnfVj018995@enl.usc.edu>

Dear class,

I already sent out emails to the students who submitted their projects
by the deadline. If you submitted your project and didn't get that
email, you need to contact me immediately.

I already sent out emails to students who requested slip day
confirming that I received those slip day requests. If you sent a slip
day request and didn't get that email from me, you need to contact me
immediately.

I already sent out emails to some students whose code I was not able
to compile on aludra. It is in your best interest to fix your code so
that it compiles on aludra and resubmit before it is late by too many
days. There might be a few more students who might get emails from us
in the next few hours if we find out that we are unable to run their
code and such. Watch out for those emails.

Thanks.

- om_p

From johnh@ISI.EDU  Wed Apr 23 22:01:01 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 23 Apr 2003 14:01:01 -0700
Subject: [Csci551-talk] Re: query regarding s7-output
In-Reply-To: <7e1bfc7e29bb.7e29bb7e1bfc@usc.edu>
Message-ID: <200304232101.h3NL11KG021685@dash.isi.edu>

On Wed, 23 Apr 2003 01:00:39 PDT, gaurav patel wrote: 
>
>
>----- Original Message -----
>From: John Heidemann <johnh@ISI.EDU>
>Date: Tuesday, April 22, 2003 10:03 pm
>Subject: [Csci551-talk] Re: query regarding s7-output
>
>> On Tue, 22 Apr 2003 21:06:33 PDT, Sumeet Savla wrote: 
>> >[1  <text/plain; iso-8859-1 (7bit)>]
>> >
>> >[2  <text/html; iso-8859-1 (7bit)>]
>> >Hi professor,
>> > 
>> >Just needed a confirmation on what i am doing...
>> >whenever i send a packet, i am logging it into the file.
>> >So, this means that if I am resending a packet, i log it again as "sending
>> >packet....".
>> > 
>> >Same happens on the receive side....
>> >i.e. if I receive the same packet again (due to an ack being dropped), i log
>> >it again as "receiving packet....".
>> > 
>> > 
>> >Sumeet
>> >
>> >
>> 
>> That sounds OK, although I'd document your choices there in the
>> README.
>
>Irrespective of choice, as per the requirements, do we HAVE to log the retansmission & duplicate messages received?
>

When you find somehting that is incompletely specified, please
document your design choice and reasoning in the README.

You have to do only what's listed in the assignment.  However, like
any specification, it will not cover all cases.  You should document
the cases that aren't covered and what choices you made.

   -John Heidemann

From johnh@ISI.EDU  Wed Apr 23 22:05:27 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Wed, 23 Apr 2003 14:05:27 -0700
Subject: [Csci551-talk] Re: Dropping acks?
In-Reply-To: <1051096204005290@lycos-europe.com>
Message-ID: <200304232105.h3NL5R6C021729@dash.isi.edu>

On Wed, 23 Apr 2003 13:10:04 BST, Sumeet Savla wrote: 
>Hi                                                            
>                                                              
>I have made all udp traffic reliable (both the joins and the  
>sends).                                                       
>                                                              
>I have a doubt regarding what should happen if an             
>acknowledgement packet is dropped.                            
>                                                              
>The timeout will occur due to lack of acknowledgement and the 
>packet will be retransmitted.                                 
>However, since we had already sent the packet which had been  
>"received" by the routers (and processed too);                
>what do we do about this resent packet - is it okay if we     
>reprocess it or should it be ignored?                         
>Are both methods okay??                                       
>I am not sure if the reprocessing leads to looping , because  
>it ddint go into loops in my test-cases!                      
>                                                              
>Or should I simply not drop acks and keep it simple?          
>                                                              
>Sumeet                                                        

THis is a common issue in real protocols, and in the real world, not
dropping ACKs is not an option.

Two common methods in Real World protocols are:

- simply reprocessing the message, assuming it does no harm
    (such messages and the operations they cause are called *idempotent*,
    meaning they can be done multiple times wihtout harm)

- keeping a cache of prior processed messages, and replying the ACK
  out of the cache

(Something to think about: which approach would be easist for
multicast subscriptions?)

I hope that your app does one of these, or that you don't encounter
this problem in the evaluation runs...

   -John Heidemann

From savla@usc.edu  Wed Apr 23 05:06:33 2003
From: savla@usc.edu (Sumeet Savla)
Date: Tue, 22 Apr 2003 21:06:33 -0700
Subject: [Csci551-talk] query regarding s7-output
Message-ID: <002c01c3094d$ba305f90$d8f97d80@Sumeet>

This is a multi-part message in MIME format.

--Boundary_(ID_+eW91y4XGLoeVKC2EZzDIQ)
Content-type: text/plain; charset=iso-8859-1
Content-transfer-encoding: 7BIT

Hi professor,

Just needed a confirmation on what i am doing...
whenever i send a packet, i am logging it into the file.
So, this means that if I am resending a packet, i log it again as "sending packet....".

Same happens on the receive side....
i.e. if I receive the same packet again (due to an ack being dropped), i log it again as "receiving packet....".


Sumeet

--Boundary_(ID_+eW91y4XGLoeVKC2EZzDIQ)
Content-type: text/html; charset=iso-8859-1
Content-transfer-encoding: 7BIT

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN">
<HTML><HEAD>
<META http-equiv=Content-Type content="text/html; charset=iso-8859-1">
<META content="MSHTML 6.00.2723.2500" name=GENERATOR>
<STYLE></STYLE>
</HEAD>
<BODY bgColor=#ffffff>
<DIV><FONT face=Arial size=2>Hi professor,</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>Just needed a confirmation on what i am 
doing...</FONT></DIV>
<DIV><FONT face=Arial size=2>whenever i send a packet, i am logging it into the 
file.</FONT></DIV>
<DIV><FONT face=Arial size=2>So, this means that if I am resending a packet, i 
log it again as </FONT><FONT face=Arial size=2>"sending 
packet....".</FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>Same happens on the receive side....</FONT></DIV>
<DIV><FONT face=Arial size=2>i.e. if I receive the same packet again (due to an 
ack being dropped), i log it again as <FONT face=Arial size=2>"receiving 
packet....".</FONT></FONT></DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2></FONT>&nbsp;</DIV>
<DIV><FONT face=Arial size=2>Sumeet</FONT></DIV></BODY></HTML>

--Boundary_(ID_+eW91y4XGLoeVKC2EZzDIQ)--

From teli@usc.edu  Thu Apr 24 10:21:18 2003
From: teli@usc.edu (parthik teli)
Date: Thu, 24 Apr 2003 02:21:18 -0700
Subject: [Csci551-talk] final exam content?
Message-ID: <d7d071d7ad2b.d7ad2bd7d071@usc.edu>

Dear TA,
   Do we have to do papers before midterm for finals ? if yes, what will be the % that will be from those papers.
  I am waiting for your kind reply :)
            thanks
                             -Parthik
======================
Hard work is the way to "Zenith". 
Faith is the 'Base' of  "Relation"


From gnawali@usc.edu  Thu Apr 24 10:14:22 2003
From: gnawali@usc.edu (Omprakash Gnawali)
Date: Thu, 24 Apr 2003 02:14:22 -0700
Subject: [Csci551-talk] Re: final exam content?
In-Reply-To: Your message of Thu, 24 Apr 2003 02:21:18 -0700
Message-ID: <200304240914.h3O9EMdP001810@enl.usc.edu>

Here is an excerpt from FAQ
(http://www.isi.edu/~johnh/WORK/CS551/SP2003/exam_faq.html)

Q: Will the final cover the whole class? 

A: Yes, but there will be a slight emphasis on the second half of class. 


you wrote:
> 
> Dear TA,
>    Do we have to do papers before midterm for finals ? if yes, what will be the % that will be from tho
 *se papers.
>   I am waiting for your kind reply :)
>             thanks
>                              -Parthik
> ======================
> Hard work is the way to "Zenith". 
> Faith is the 'Base' of  "Relation"

From johnh@ISI.EDU  Thu Apr 24 16:54:35 2003
From: johnh@ISI.EDU (John Heidemann)
Date: Thu, 24 Apr 2003 08:54:35 -0700
Subject: [Csci551-talk] query regarding s7-output
In-Reply-To: <002c01c3094d$ba305f90$d8f97d80@Sumeet>
Message-ID: <200304241554.h3OFsaEm003084@dash.isi.edu>

On Tue, 22 Apr 2003 21:06:33 PDT, Sumeet Savla wrote: 
>Hi professor,
> 
>Just needed a confirmation on what i am doing...
>whenever i send a packet, i am logging it into the file.
>So, this means that if I am resending a packet, i log it again as "sending
>packet....".
> 
>Same happens on the receive side....
>i.e. if I receive the same packet again (due to an ack being dropped), i log
>it again as "receiving packet....".
> 
> 
>Sumeet
>
>

Exactly how logging and retranmissions interact is not defined by the
Project B assignment, so you should pick a reasonable way to do it and
document that in your README file.

What you describe sounds reasonable.
   -John Heidemann