uneven log balancing and elevated processing time after node reboot
facing an interesting issue
version - BIG-IP 14.1.4.6 Build 0.0.8 Point Release 6
iapp > pool > i7 rule > if traffic matches i7 uri rules then it's redirect to "subpool"
"subpool" consists of 2 node/member f5 pool with 120 minute persistance, oneconnect multiplex and least connections lb with 2 linux virtual servers behind it - can provide more detail lmk what's required
issue is whenever one member server in the "subpool" is rebooted, often during slow periods during maintenance window (usually by automated process) application processing time on it goes from a few milliseconds to 2-3 seconds and it hardly gets any connections, while the server that has not been rebooted processes normally while taking most of the connections; the only workaround we have so far is rebooting both servers at once; the issue does not self-remediate with time, we waited about a day (24 hours) to see if it would, I have yet to try and modify the persistance profile/timeout etc since this is happening in a production environment and I can't change stuff like that on the fly while the issue is occuring
has anyone seen an issue like this? what's the best way of going about troubleshooting?
we suspect the persistance rule, aside from possibly deleting all the persistance records when the issue is happening how can I determine it's the cause or not?
Thanks!