I have two BGP servers on OpenBSD using bgpd that are located in the same DC; I tried to enable multihoming for the same networks on both of the servers, but upon doing so all hosts in that DC started experiencing high latencies; this lasted for about 20 minutes until I reverted the changes making a single box the main bgdp server. There were no issues on either of the BGP servers and both were in Established state. Could someone explain what was going on with the high latency and why that might've occurred?
Asked
Active
Viewed 127 times
0
1 Answers
0
What most likely happened is you created a loop that used the multihomed provider as a transit AS and your packets went back and forth between BGP servers until they were finally routed.
It's hard to say for sure without seeing the exact prefixes your were advertising to the multihomed peer.
Alo
- 240
- 1
- 7