Allow configuration of probes for created containers in Helm chart

[djessich]

Is your feature request related to a problem? Please describe.
Probes cannot be configured when deployed with Helm, although each container should have a liveness, readyness and startup probe configured to conform to Kubernetes best practices.

Describe the solution you'd like
Please allow configuration of probes and provide a reasonable default. This should be available for all created containers.

Describe alternatives you've considered
Currently there is no possibility to configure probes.

Additional context
Each container should have a liveness, readiness and startup probe configured to conform to current Kubernetes best practices. Currently this chart does not support this feature.

[cniackz]

Synced with Allan on this, and we discovered that liveness and readiness is already implemented, all you need to do is add that into the values:

File: operator/helm/tenant/values.yaml
Content:

  # Liveness Probe for container liveness. Container will be restarted if the probe fails.
  # Refer https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes.
  liveness:
      tcpSocket:
        port: 9000
      initialDelaySeconds: 15
      periodSeconds: 20
  # Readiness Probe for container readiness. Container will be removed from service endpoints if the probe fails.
  # Refer https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
  readiness:
      tcpSocket:
        port: 9000
      initialDelaySeconds: 5
      periodSeconds: 10

And then call it:

helm install --namespace tenant-ns \
  --create-namespace tenant minio/tenant \
  -f operator/helm/tenant/values.yaml

That way containers will have them:

      livenessProbe:
        tcpSocket:
          port: 9000
        initialDelaySeconds: 15
        timeoutSeconds: 1
        periodSeconds: 20
        successThreshold: 1
        failureThreshold: 3
      readinessProbe:
        tcpSocket:
          port: 9000
        initialDelaySeconds: 5
        timeoutSeconds: 1
        periodSeconds: 10
        successThreshold: 1
        failureThreshold: 3

But we are missing startup probe and Allan will create PR for it soon

[allanrogerr]

Startup probe has been implemented as part of the Helm install. This functionality will be available in the next release. This probe may be added as part of operator/helm/tenant/values.yaml. See example value content below:

  # Startup Probe for container startup. Container will be restarted if the probe fails.
  # Refer https://kubernetes.io/docs/tasks/configure-pod-container/configure-liveness-readiness-startup-probes/
  startup:
    tcpSocket:
      port: 9000
    initialDelaySeconds: 15
    timeoutSeconds: 10
    periodSeconds: 20
    successThreshold: 1
    failureThreshold: 2

After a helm install with the above example startup probe, the containers will register:

startupProbe:
      failureThreshold: 2
      initialDelaySeconds: 15
      periodSeconds: 20
      successThreshold: 1
      tcpSocket:
        port: 9000
      timeoutSeconds: 10

[allanrogerr]

fixes: #1370