file CAN_INSTALL is missing from your config directory

[nextcloud]

file CAN_INSTALL is missing from your config directory

[wikoion]

The startupProbe is an inconsistent solution. The delay will always need to be longer than it takes for the database to come up, if this takes longer than usual, because your cluster is waiting to claim resources for example, this will still fail. The initContainer solution actually checks if your container is ready. You can test my PR to see if it fixes the issue for you?

[xyhhx]

If anybody is still getting this, it's likely due to the same issue (installation being interrupted).

In my case, it helped to increase the settings for the readiness and liveness probes:

    startupProbe:
      enabled: true
      periodSeconds: 30
      failureThreshold: 10
    livenessProbe:
      enabled: true
      initialDelaySeconds: 60
      periodSeconds: 30
      failureThreshold: 10

Also, if you're using an external database, and are managing it and this chart with kustomize, you can set the nextcloud kustomization to wait for the postgress customization using dependsOn. For example:

---
apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: nextcloud-postgresql-release
  namespace: flux-system
spec:
  path: ./kubernetes/workloads/nextcloud/postgresql/release
  prune: true
  sourceRef:
    kind: GitRepository
    name: flux-system
  wait: true
  interval: 30m
  retryInterval: 1m
  timeout: 5m

---
apiVersion: kustomize.toolkit.fluxcd.io/v1
kind: Kustomization
metadata:
  name: nextcloud-release
  namespace: flux-system
spec:
  path: ./kubernetes/workloads/nextcloud/nextcloud/release
  prune: true
  dependsOn:
    - name: nextcloud-postgresql-release
  sourceRef:
    kind: GitRepository
    name: flux-system
  wait: true
  interval: 30m
  retryInterval: 1m
  timeout: 15m

[niklashagman]

This issue ticket is probably related to #52, #49 and #42.
All because Nextcloud is not failing correctly when database is not yet available. See my suggested solutions above.

[niklashagman]

Recently StartProbe was added to Nextcloud helm charts.

startupProbe:
  enabled: true

This does not solve the initial problem but it makes Nextcloud fail correctly in my case. Meaning after a second retry the database is now up and nextcloud will successfully install itself understanding that it failed the first try and will not ask for CAN_INSTALL file.

[tcardonne]

Even with the startupProbe enabled, my Nextcloud deployment asks for the CAN_INSTALL file after failing because the included database is not ready on the first try.

Am I missing something or only something like #67 will fix this issue ?