This message was deleted.

We also have this issue and have started looking at etcd as a potential source. Did you resolve your issue?

We resolved by tearing down the cluster and rebuilding unfortunately. Our issue looked more like a CNI driver issue than an etcd one, honestly