From: Petr Lapukhov (petr@internetworkexpert.com)
Date: Sun Jun 25 2006 - 14:39:09 ART
I don't see a reason to put RIP2 here...:)
First, to begin with, you should consider you addressing scheme.
If you have clear hierarchy, with 2 or 3 layers, that's fine. If not,
you should consider readdressing, since uneffective addressing
plan really hurts scalability, no matter what routing protocol you use.
Next, to OSPF. It's not that really CPU consuming nowdays, if
you have network design with scalable addressing. Well, sure,
if you have 2500 and 1600 routers, it's not a best idea to run OSPF
on them ;)
Now, to say somehting about RIP. Nowdays, you don't have real
problems with hop-count metric. It's rather RIP's slow convergence,
and loop-prone behavior, that may hurt.
So that's what I would recommend:
1) Make sure you have scalable addressing plan for you network
2) Run OSPF with Area 0 in core, and put reginal offices into stub areas.
In some simple cases you may consider static routing (reliable static
routing, e.g.), or even ODR for regional offices.
3) Summarize regional networks, while injecting them in backbone
4) If you need redundant link for regional offices, send them two or more
defaults,with different costs, and injects summaries into backbone
accordingly.
5) Tune OSPF timers to speed up convergence, and fault recovery (OSPF may
benefit from 1s dead timer, for instance)
HTH
-- Petr Lapukhov, CCIE #16379 petr@internetworkexpert.comInternetwork Expert, Inc. http://www.InternetworkExpert.com <http://www.internetworkexpert.com/> Toll Free: 877-224-8987 Outside US: 775-826-4344
2006/6/25, Radioactive Frog <pbhatkoti@gmail.com>: > > Hi Group, > > A company have 4 offices in metropolitan city. All 4 are connected with > high > speed network (1 or 2 T1's ). These four offices are in fully mess. Each > metropolitan city has approximately 40-60 regional offices connected to > it. > They all have Leased link say 64 or 128K. > > Total: 4 metropolitan city offices x 40-60 offices in each > > One of the metropolitan city is Headquarter and have database and mail > server inside. That got to be accessed from the all remote sites (regional > and other metropolitan cities). > > Which protocol is best for this scenario ?? RIPV2 or OSPF ? (no bgp, eigrp > or igrp). > > From my calculation: RIP should be enough as no of hops are less then the > required. > ------------------------------ > > Headquarter--------->4 Metro offices(4hops) ------->50-60 remote offices > (1 > hop) = total 6 hops total > > So in worse case if packet goes through messed network (4 metro > offices) we > will be not using more then 6 hops. > In addition to this RIP-V2 doesn't need more processing power and high end > router. > OSPF requires memory and high end routers. > > What you guys think ?? > > _______________________________________________________________________ > Subscription information may be found at: > http://www.groupstudy.com/list/CCIELab.html >
-- Petr Lapukhov, CCIE #16379 petr@internetworkexpert.com
Internetwork Expert, Inc. http://www.InternetworkExpert.com Toll Free: 877-224-8987 Outside US: 775-826-4344
This archive was generated by hypermail 2.1.4 : Sat Jul 01 2006 - 07:57:33 ART