So I have a question regarding sparse-mode multicast using auto-rp to determine the rendevous point. In my network we have four sites: our primary site in Chicago where the data center resides, a secondary site in the next town over, which is connected via EIGRP (we have our own fiber between these two sites) and then a site in the UK and NY. For redundancy, I wanted to enable at both the Chicago and the secondary site with ip pim send-rp-discovery and ip pim send-rp announce, with the Chicago site having a higher IP address and thus the preferred RP in an election.
For the purpose of this question we'll say, Chicago is using 10.6.6.6 and the secondary site 10.5.5.5 on their respective loopback interfaces).
I would expect to see when I do a show ip pim rp mapping that Chicago is the only RP. But I show both subnets having an RP.
Why is this, if highest IP wins?