Daniel Finneran
Daniel Finneran
Can you explain what your hardware is? this largely looks like the backing `etcd` is unstable.
Yes, it's more than likely that `etcd` is having issues, thus causing both the API server and leaderElection issues.
Cool, thanks for clarifying the underlying network problems 👍
I think the annotations per loadBalancer makes sense.. I'll add this to the roadmap!
Annotations per loadBalancer were added in #819
Apologies.. Being a bit slow from the Christmas break.. There should be no issue just recreating the manifests.. the leaderElection code will handle the changing of the underlying kube-vip pods.
Given the throttling, it looks like something is spawning multiple watchers on the same leases.
Fix incoming, along with a MASSIVE set of end to end testing so this won't happen again.
https://github.com/kube-vip/kube-vip/pull/597
This is an old conversation, please feel free to re-open or create a new issue if needed.