Lex Rivera

ATM i'm cleaned all my nodes from leftover files, so i can't give you exact configuration. I used default storageClass with 3 replicas. All nodes are schedulable. Node with attached...

Can you try non-clean shutdown, simulating power outage or something like this? Killing VM? I think this is what caused leftover files, but i'm not sure.

Two options: "You don't need to delete the dangling vol_data.json in the old mountpoint directory by manual. After restart of the longhorn-csi plugin automatically and waiting for several minutes, the...

>@rlex how can I know a volume is not "live" volume. ? I think if it attempts to cleanup it's not live, but you can check volume details in longhorn...

Unification of loki, mimir and tempo charts will be great indeed. First thing that comes to mind is different names of gateways in charts - nginx, gateway, nginx-gateway... > How...

Faced same issue with vector helm chart. Mainly this part: ``` labels: forwarder: "vector" file: '{{ file }}' stream: "{{ stream }}" source_type: "{{ source_type }}" k8s_pod_namespace: "{{ kubernetes.pod_namespace }}"...

I'm gonna assume kube-state-metrics should be scraped by prometheus specified in coroot, so KSM should either have servicemonitor or annotations, depending on your prometheus setup

using minio by any chance?

Oh, good to know - i can disable it without loss of any functionality? Looks like chart have queryFrontend.query.enabled parameter which is true by default

good idea, that will also allow whole tempo stack to run with readonly root fs (with linked PR)