Cisco Support Community
Showing results for 
Search instead for 
Did you mean: 
Community Member

CE - PE routing protocol choice? BGP or RIPV2?

what do you think would be the best choice for a CE - PE routing protocol, for a site with redundant links to redundant PE routers. the service provider will only offer static, BGP, or RIPV2 with standard timers configured. from you experiences, what do you think would provide the best possible failover / reconvergence time?

Thanks in advance.

Hall of Fame Super Gold

Re: CE - PE routing protocol choice? BGP or RIPV2?


Knowing what kind of connection it was from the PE to the provider might help determine my choice. But my thoughts are like this:

- if there are redundant links from the customer to the provider then I would prefer a dynamic routing protocol over static routes. The dynamic protocol will be more effective in detecting link failure than static routes would be and in providing failover to the surviving link.

- given a choice between RIPv2 or BGP I might slightly prefer RIPv2, based primarily on the more simple configuration and more easy troubleshooting if something does not work as expected.

I expect that some of my colleagues may suggest that BGP is better between a customer and a provider. But my sense is that BGP is primarily implemented for its support of policy in routing decisions. And I do not detect much requirement for policy in your description of your situation.



Re: CE - PE routing protocol choice? BGP or RIPV2?

I would go with BGP even if only for the reason its NOT rip

rip sends data every few seconds as opposed to only on link state change

if your network is somewhat large and your not multi-homing then I'd do BGP on the edge router and Ospf inside


Re: CE - PE routing protocol choice? BGP or RIPV2?

What are "standard" timers?

Hall of Fame Super Silver

Re: CE - PE routing protocol choice? BGP or RIPV2?

Hello Andrew,

if your CE router has direct links to the PE routers (no devices in the middle) you can use BGP with ebgp fast-external fallover in this way BGP convergence is the time to detect that the link is down and not the BGP hold time (20-30 seconds in normal links, hundreds of ms on STM/SDH links).

use maximum-paths N where N is the number of PE-CE links.

Without using this option RIPv2 is a little faster but it will take the time to invalid the routes received on the failed link.

I would use BGP + fast external fallover + maximum-paths N

If your links are not direct with default timers RIPv2 is faster.

Hope to help


CreatePlease to create content