Joaquín Muleiro

+1 I'm experiencing the same issue running Loki and Grafana in Kubernetes using the official Helm charts. In case it helps, the versions I'm running are: **Loki** - 2.6.1 **Loki...

My bad here, but I never got to set up those pages you're describing. Pretty sure I'm missing some configuration, since if I access my Loki gateway's /ring or /distributor/ring,...

Yes, I'm running it in Kubernetes. I will try the port-forward approach first, thank you! :)

Okay, this made me feel dumb. In case anyone is wondering, or needs to know in the future: The "too many unhealthy instances in the ring" log will also show...

@QuentinM-Hilbtec I could be mistaken, but as I understand it, if using `read` and `write` instead of `singleBinary`, Loki will return "too many unhealthy replicas" if either the read or...

Hello there! As per my comment in #6231, this appears to still be happening. Any idea what it could be about?

+1 Just verified that it happens with some non-base images, like nginx, where the default user is root.

> so to install the chart from that specific repo I have to prefix my chart name with a repo name, `hashicorp/consul`, while Terraform Helm provider does not need that...

I'm also seeing this error sporadically. It occurred in one out of 25 identical `ubuntu-20.04` containers created from a matrix strategy.

Agree with @Neurobion, I set up alerts recently for Velero backups, just to find out that they don't work as expected. We need a fix, as these metrics are not...