So I know in VPNv4 peerings the next-hop (in iBGP) is automatically set. I also know that if you throw in RRs that the code is smart enough to know the RR clients shouldn't receive VPNv4 routes with their NHs changed. Obviously we don't want traffic going through the RRs. Now here is what I don't get.
If I have 2 RRs (redundancy yay) and they peer only via VPNv4 (regular iBGP VPNv4) why is their NH not modified when they share routes? I know it's working out of the box because I didn't need to do it in my lab. I just want to know the mechanisms as to why or how they know they are RRs and between each other. Is it because they are passing routes with originator IDs and they do some type of internal check (wild guess)?
Also, bonus question :D I've found some documentation state that the RR should have LDP enabled but I've found that you don't need it as long as the RR is out of the dataplane/path. Can someone confirm ?
No comments:
Post a Comment