Thomas Jungblut

Lots of things to unpack, it sounds very similar to #401 to a degree. Can you create a [CPU profile with pprof](https://pkg.go.dev/runtime/pprof) for us please? Maybe that sheds the light...

Thanks for all the information. > with the default 4K pagesize it constantly splits inodes and i guess the heavy writes are coming from this? it certainly seems the splitting...

Glad you've got it to work fast :muscle: Anything that would need to make its way back into bbolt?

To repro this add a `time.Sleep(11 * time.Millisecond)` sleep before this line: https://github.com/etcd-io/etcd/blob/4ece5561d4ba75563fd15ca2ebb4c6f572c1431d/server/etcdserver/api/v3compactor/periodic.go#L125 The recorder (backed by an unbuffered channel) is supposed to act as a semaphore here, but the...

I'll try to repro this over the course of today, can you please give us the full config/helm chart for repro? Or at least the YAMLs of the respective STS/service...

Sorry @rahulbapumore but this is far away from being reproducible for me. What are all of those pieces? ``` containers: - args: - -- - /bin/bash - -c - /opt/redirect/stdout-redirect...

Okay, I believe we're conflating many different issues here now. > same as having N member etcd cluster and out of N more than N/2 members are down. That's the...

/test e2e-aws-etcd-recovery

> This is problematic when you run etcdctl with cron and you want to receive a mail from the cronjob only when an error occured. I agree with you on...

We see such things very often when operators misbehave: ``` flat flat% sum% cum cum% 2256.76MB 57.10% 57.10% 2256.76MB 57.10% bytes.makeSlice 1199.99MB 30.36% 87.46% 1199.99MB 30.36% go.etcd.io/etcd/raft/raftpb.(*Entry).Unmarshal 476.06MB 12.04% 99.50%...