i have been studying about fast
rerouting for eigrp,
but i have beens tuck on understanding some parts
of it.when i tried to read the cisco document regarding this i couldn't able to catch the points.
1)why per-prefix based computation is better than per-link based computation?
2)I can not able to understand the working of the following tie-breakers
i)interface-disjoint
ii)Linecard-disjoint
iii)lowest-repair-path-metric
iv)shared risk link group
3)why load sharing should be disabled in order to influence the tie-breaker ?
can some one explain me ?