Remove leftover Grafana/Prometheus artifacts after helm uninstall #46

New issue

Open

opened 2026-08-02 12:56:27 +00:00 by lab · 0 comments

lab commented

2026-08-02 12:56:27 +00:00

Owner

Context

Prometheus stack was uninstalled via helm uninstall prometheus -n monitoring (helm-chart based kube-prometheus-stack, repo config removed, caddy grafana route removed, monitoring/README.md rewritten to beszel-only).

Leftovers still in cluster (helm does NOT track these)

CRDs from kube-prometheus-stack: monitoring.coreos.com group (alertmanagers, prometheuses, prometheusrules, servicemonitors, probes, thanosrulers).
Dead secrets: grafana-admin-credentials, prometheus-kube-prometheus-admission in monitoring namespace.

TODO

Delete the monitoring.coreos.com CRDs (verify nothing else uses them first).
Delete the two dead secrets.
Optionally check for orphaned longhorn volumes / configmaps from grafana/prometheus.

## Context Prometheus stack was uninstalled via `helm uninstall prometheus -n monitoring` (helm-chart based kube-prometheus-stack, repo config removed, caddy grafana route removed, `monitoring/README.md` rewritten to beszel-only). ## Leftovers still in cluster (helm does NOT track these) - CRDs from kube-prometheus-stack: `monitoring.coreos.com` group (alertmanagers, prometheuses, prometheusrules, servicemonitors, probes, thanosrulers). - Dead secrets: `grafana-admin-credentials`, `prometheus-kube-prometheus-admission` in `monitoring` namespace. ## TODO - [ ] Delete the monitoring.coreos.com CRDs (verify nothing else uses them first). - [ ] Delete the two dead secrets. - [ ] Optionally check for orphaned longhorn volumes / configmaps from grafana/prometheus.