Ovidijus Balkauskas

We're also affected by this issue. Additionally, `salt-ssh` did not respect our master's conf option `git_pillar_provider: gitpython`, reverting to the default `pygit2` preference for the 2nd pillar/state render (until we...

Agree, this is a very useful enhancement.

[I've submitted a PR](https://github.com/k8up-io/k8up/pull/948) that adds a configuration option that is somewhat related to this issue. It adds an operator-scope boolean variable `BACKUP_SKIP_WITHOUT_ANNOTATION` that, when changed from the default `false`...

Having the same issue here, but with Everspace.

We're not sure what scenarios cause it, as we only observe it in production (where end-users create and join video calls). At the very least I can confirm the issue...

I'm still in the process of gathering a reproducable case, however one behavior I immediately observe is that the metrics endpoint and livekit-cli definitely has a delay/latency to each other:...

While we were unable to reproduce the issue in an isolated environment, we still observe erratic metric behavior. For example, one Livekit node now reports that the published video track...

Same situation with `livekit_participant_total` on livekit-server v1.6.0: ``` # curl -sq http://10.0.1.13:6789/metrics | grep livekit_participant_total livekit_participant_total{node_id="ND_TdXj9kR3xXX3",node_type="SERVER"} -1740 # curl -sq http://10.0.1.14:6789/metrics | grep livekit_participant_total livekit_participant_total{node_id="ND_9ChQrBP8u99f",node_type="SERVER"} -2063 ```

We'll update to 1.6.1 (that has #2672) and monitor if this issue is gone

I can confirm that since 1.6.1 this issue has been resolved. Closing.