IBM Operator for Apache Flink fails to automatically update the image to 1.1.8 after upgrading

Symptoms

After you successfully upgraded to Event Processing 1.1.8, the Flink images fail to automatically update to 1.1.8 and display the images of earlier versions.

Causes

Event Processing fails to trigger the mutating admission webhook that changes the image versions.

Resolving the problem

To resolve the problem, trigger the mutating admission webhook to update the image version to 1.1.8 by running the following command:

kubectl patch flinkdeployment <instance-name -n <namespace> \
  --type='json' \
  -p='[{"op": "replace", "path": "/spec/flinkConfiguration/table.exec.source.idle-timeout", "value":"31 s"}]'

The previous command changes an arbitrary Flink configuration parameter in the FlinkDeployment custom resource to trigger the mutating admission webhook.