The primary went down, the backup took over thanks to keepalived, but it
immediately went down as well.
kdb appeared on both machines briefly, I saw that the coredump was in
the netfilter code. As I tried to figure out how to get the entier
backtrace ("bt | more?") both machines locked up. I can't find anything
in the logs.
It was experiencing high load.
It looks like the bug happened due to the traffic given that both
machines locked up.
Are there any known bugs in IPVS shipped w/ 2.4.23 which could cause
this, if so which ones? I would like to justify replacing the commercial
load balancer but I can only do this if I can explain why LVS failed.
|