Uninstall
Remove Harthelm, and what stays behind.
Uninstalling deletes the bundled database
helm uninstall deletes the platform database of the bundled mode. Dump it first if you might come back.
Delete the teams
Workspaces, their volumes and users' databases live in the team namespaces rg-*. They are core's objects, not Helm's,
so Helm does not remove them. Delete the teams in the console while core still runs (each team's Settings →
Delete team…), or delete the namespaces yourself after the uninstall:
kubectl delete namespace -l harthelm.dev/managed=trueKeep a last copy
Run a dump and keep the master key, in case you need the data later:
kubectl -n harthelm-system create job --from=cronjob/harthelm-backup harthelm-final-dump
kubectl -n harthelm-system get secret harthelm-master-key -o yaml > harthelm-master-key.yamlUninstall the release
helm uninstall harthelm -n harthelm-systemUnder Argo CD, delete the Application.
Remove what stays
| Left behind | Remove with |
|---|---|
The backups volume harthelm-backups (kept on purpose) | kubectl -n harthelm-system delete pvc harthelm-backups |
Under Argo CD, the database Cluster (marked not to delete) | kubectl -n harthelm-system delete cluster.postgresql.cnpg.io harthelm-db |
| Secrets you created (OIDC client, master key, SMTP login) and the ones core generated | kubectl delete namespace harthelm-system |
Cluster-scoped objects with the label harthelm.dev/managed=true (Cilium backstops) | kubectl delete ciliumclusterwidenetworkpolicies -l harthelm.dev/managed=true |
| cert-manager's certificate Secrets of custom domains | deleted with the namespace |
The chart's own cluster-scoped objects (ClusterRoles, bindings, the admission policy) go with the release. The operators (cert-manager, CloudNativePG) are yours and stay.