[c-nsp] Difference in IP FRR Link vs Per Prefix Option

[Adam Vitkovsky]

> So it means with Per link we can get node protection in some cased means
it will see both Inequality 1 & 3 if it gets 3 good enough otherwise 1 is a
must. 
Yes I believe per-link LFA is concerned only with inequality 1, if the
chosen LFA node happens to have a better path to destination other than via
primary next-hop than you'd get the advantage of node protection. 


> With Per Link router will do SPF computation for all the prefixes learned
through Primary next-hop so in terms of prefix coverage is concerned i don't
see any prefix is left for LFA candidate only point is with per link is that
all the prefixes with LFA will share the same backup path. 

Well yes both flavors (per-link/prefix) will try to find a backup path for
all prefixes in the link state database. 
How I understood it is that per-link LFA will not run a full spf calculation
which results in less probability of finding an LFA candidate and in
addition to that it will select the first available LFA candidate and stop
the calculation afterwards (which is fine since it's not concerned with
inequality 3 anyways, so whatever it finds first is fine, according to the
limited criteria). 


> In IOS-XR 4.3.1 Cisco introduced rLFA (remote LFA) which when used in
conjunction with LDP can greatly enhance coverage by bypassing spots where
traffic may loop.  
Yes I'd definitely recommend rLFA it's the best thing we have currently,
however it runs on a fairly new code. 
We did use te-tunnles as backup paths to get the rLFA behavior manually and
it did provide full coverage in ring topologies. 
But that was prior to migration to safe harbor 4.2.3 as it introduced a bug
on this functionality. 


adam