[email protected]
[Top] [All Lists]

Re: Motivation of IPFRR

Subject: Re: Motivation of IPFRR
From: "Alia Atlas"
Date: Tue, 5 Aug 2008 11:09:23 -0400
Hi Andras,

One advantage that I see for an IP-FRR solution is that it enables a solution that offers end-to-end protection, regardless of multiple routing areas.  It is possible to protect against the failure of an ABR, for instance. 

Another thing is that while we talk about IP-FRR, it works just as well for LDP.  There are a number of MPLS networks that are running just LDP and do not need/want the complication of running RSVP as well on top just for FRR.

LFA, in particular, is much lower for operational complexity than RSVP-TE FRR.

Hopefully this gives you some ideas,


2008/8/5 András Császár <[email protected]>
Hi All!

I really like the idea of having an FRR solution for native IP networks and so I have been following the evolution of IPFRR solutions for quite some time now (LFA, not-via, MRC, FIR & co.), and I think I know quite well what are the different proposals about.

Yet, the other day I was asked a question, for which I couldn't and still can't give a profound answer. The question was something like this:

Why do we need a separate IP-FRR solution, why is it not enough to use an MPLS detour to a (next-)next-hop to protect links?

The motivation behind it was that basically all major router vendors (C*, J*, R*) support it, why do we need a dedicated native IP solution?

Up to know I thought that there is MPLS-FRR for MPLS networks, there is IP-FRR for IP networks. But now I think that node/link protection of MPLS-FRR is straightforward to apply to IP networks as well. I mean, you don't have to bother about managing a full fledged MPLS network, although there will be some protection LSPs for FRR purposes but these are quite automatic.

Are there scenarios out there which require the extreme high resilience provided by FRR but which do not afford having even a little MPLS only for protection? Under normal conditions, the network is basically still a pure IP network...

I went through the intro of all major IPFRR publications out there, but each motivated IPFRR with the need of an IP-only solution, but what is the motivation for having a native IP-only solution?

Do I miss something, or is the answer only something like management-overhead of MPLS?

Any answers would very appreciated!

rtgwg mailing list
[email protected]

rtgwg mailing list
[email protected]
<Prev in Thread] Current Thread [Next in Thread>