Setting livenessprobe in satellite config fails

[Jeremy643]

I'm trying to update livenessprobe for the linstor-satellite container but it's not being patched.

For example, my linstor satellite config has this liveness probe set;

       livenessProbe:
          exec:
            command:
            - /bin/bash
            - -c
            - /entrypoint/liveness-script.sh
          failureThreshold: 3
          periodSeconds: 10
          successThreshold: 1
          timeoutSeconds: 1

but the satellite pods are still showing;

   livenessProbe:
      failureThreshold: 3
      periodSeconds: 10
      successThreshold: 1
      tcpSocket:
        port: linstor
      timeoutSeconds: 1

[WanzenBug]

Untested, but perhaps you also need to add

tcpSocket:
  $patch: delete

Not sure how the API handles multiple probe types. I believe it causes errors. You can check the status output of the LinstorSatellite resource for details

[Jeremy643]

Thanks, that seems to have worked. I took a brief look at the code and how the LINSTOR satellite configuration resources get merged, would we not expect the user defined liveness probe to replace the tcpSocket one?

[WanzenBug]

We merge stuff using kustomize. And that mostly does yaml merging by recursively merging things.