Lucas Servén Marín

hmm interesting, I don't understand why we have thrashing. Kilo only recreates resources when it's own rules change, not when any other process's resources change. Do you have any more...

hi @tuapuikia yes, you can use the internal IP as the endpoint for the WireGuard. To do this, set the `kilo.squat.ai/force-endpoint` annotation (https://kilo.squat.ai/docs/annotations#force-endpoint) on the nodes that should be the...

Hi @faraonc that _is_ very high indeed and not normal for Kilo! ~The Kilo pods expose profile endpoints at~ I'm surprised to learn that I didn't yet add profiling endpoints...

When we've seen high CPU issues in the past, it's tended to be either due to constant reconciliation because some configuration doesn't match or Kilo and another process are fighting...

Hi @faraonc i made the PR earlier today: https://github.com/squat/kilo/pull/305 Hope to merge today and you can test out. Otherwise You can built it yourself and test :) Something funny is...

@faraonc the PR just merged, so you should be able to use the `squat/kilo:0ddeea3d7840acc7d6c8f021310e9b39e7dad6f5` tag in a few minutes to test out the pprof endpoint. In any case it seems...

As i mentioned earlier, i strongly suspect that Kilo is getting confused with the WireGuard configuration. Any info you could share about the peers in the cluster / what other...

Hi @faraonc, any update?

Totally understand 👍 Let me know if you need help debugging. You can also find me in the #kilo channel in the k8s slack.

Thanks for the detailed report @3rmack. It occurs to me that there might be an issue with the default policy of the `FORWARD` chain on the `filter` table in iptables...