[ns] MNS 2.0 and queueing disciplines
fox.nar at tiscali.it
Fri May 21 08:02:29 PDT 2004
Dear NS users and Bo Peng
you are right, te_db is used to take trace of the resource avaiability during
simulation. Trigger update
has some strange behaviour. I had to change in "routing-nochange" the time
to recall the function. According to the simulation that I'm running, I
get an error since routing-nochange is called at a negative value of $time.
I'm having some problems, too. My work is centred on the rerouing mechanisms.
As a matter of fact I can't use all queueing disciplines. Droptail seems
to be the only one working. CBQ gives error when you tear down and up a
link but the problem doesn't rely on MNS 2.0.
DRR, RED, SFQ and FQ work fine only with some rerouting mechanism. and
I don't understand why!!! Have you got any idea regarding it?
I look forward to hearing from you soon,
>I am doing simulation about routing in MPLS/IP network by MNS2.0. What
>trying to do is to find out the mechanism of control part of routing and
>evaluate the performance. I use the constraint-routing approach and DV
>algorithm. I have got some good results and seen clearly the route is chosen
>the available resource and dynamic re-compute the new route when resource
>changed by release the path.But there is still something I do not understand,
>like the trigger update mechanism and frequence and the relationship between
>the TE_DB(I have read the the MNS code and it seems use this as the resource
>So if someone is also doing the MPLS simulation or interested in this,
>can talk about this and try to finger something out together. I think we
>benifit each other by discussing it.
>I am looking forward to hear from you soon.
Tiscali ADSL libera la velocita'!
Attiva Senza Canone entro il 31 maggio: navighi a 1,5 euro l'ora per i primi
3 mesi,se scegli il modem e' tuo in comodato gratuito e in piu' hai gratis
SuperMail per 12 mesi. Non aspettare, attivala subito!
More information about the Ns-users