OSPF vs. RIP-2

[Harmont12345]

The main difference between the two is convergence time. RIP-2 considers link as failed after 180 seconds not receiving updates from it, compared to 40 seconds for OSPF. And OSPF sends route updates immediately in contrast to RIP which sends updates every 30 seconds.

Since "link failed" timers are adjustable, does that mean RIP would converge as fast as OSPF if it would simply send route changes immediately?

--- Nikita Andreev       

[Harmont12345]

It's becoming more interesting. As it turned out RIP does send immediate route update on link failure, it's called "Triggered Update".

--- Nikita Andreev

[Alessio Andreoli · Accepted solution]

Hi Nikita,

i like your observations.. it nearly seems that RIP is more modern that OSPF !!! I guess that the "important" or "main" differences between the two are not limited to the convergene time... I would definitely remember among these differences:

1) Metric (OSPF link cost vs RIP hop count)

2) Theinteraction with BGP

3) The flexibility with MPLS

4) OSPF is a fully loop-free portocol

5) OSPF is classless by nature vs the classful nature of RIP (even if RIPv2 can manage classless routing)

6) OSPF granularity (many different LSA packets with consequent possibility to affect routing in many ways)

7) Bandwidth consumption

8) Network diameter. OSPF can cover (according to best practices) up to 90 router per area and up to 1000 routers per domain (divided in areas). RIP has massive problems in scalability.

9) Stability

and quite a lot more...

Having said this, in 2012 you can still find companies working with RIPv2!!! And by the wa it is worth to understand it!

Back to your question, OSPF converges in a very shorter time than RIP because it works on a "per link" basis. As a link goes down the OSPF topology is recomputed (common design allows only an OSPF area updated).

Take Care

Alessio

[Harmont12345]

Well, I know about link cost and bandwidth consumption. Scalability is not an issue for our network, we are far from 90 routers.

But your answer definitely makes me think that my view is too narrow. I haven't even read about BGP and MPLS. Still preparing for CCNA. Thanks for the good answer, Alessio.

By the way, do you have an idea how heavy in fact RIP is. It seems to me, that with modern link speeds RIPs bandwidth consumption is unnoticeable.

--- Nikita Andreev