Can someone tell me best practice to have a redundant RP in PIM. We need it to failover fairly quickly.Also does one PIM group work over separate ospf area?
I would think that it would because PIM is its own thing - the big thing is to maintain valid reverse-path reachability, so basically as long as you’re not filtering your routes it should work okay - basically PIM tries to route multicast packets away from the unicast source IP, so as long as that address is correct in your routing table (not sure how it would behave in a NSSA or STUB area).
What would you recommend for RP redundancy?
Really not sure about RP redundancy (my direct experience with multicast is pretty limited) - maybe if the RP address was a VRRP address of a pair of routers, that would work, since the VRRP would be the same for everyone. I’d thought about an anycasted IP as the RP address, but that wouldn’t work because clients using one RP would not get streams from a source that reached a different anycasted copy of the RP…