Gyula Fora created FLINK-37430:
----------------------------------
Summary: Operator hides the actual error on deployment issues
Key: FLINK-37430
URL: https://issues.apache.org/jira/browse/FLINK-37430
Project: Flink
Issue Type: New Feature
Components: Kubernetes Operator
Reporter: Gyula Fora
The operator often hides the actual issue for deployment errors and do not
correctly report the error events:
{noformat}
>>> Event[Job] | Warning | CLUSTERDEPLOYMENTEXCEPTION | Could not create
>>> Kubernetes cluster "flink-state-benchmark".{noformat}
Whereas the actual error is visible in the status:
{noformat}
>>> Status[Job] | Error | FAILED |
>>> {"type":"org.apache.flink.kubernetes.operator.exception.ReconciliationException","message":"org.apache.flink.client.deployment.ClusterDeploymentException:
>>> Could not create Kubernetes cluster
>>> \"flink-state-benchmark\".","additionalMetadata":{},"throwableList":[{"type":"org.apache.flink.client.deployment.ClusterDeploymentException","message":"Could
>>> not create Kubernetes cluster
>>> \"flink-state-benchmark\".","additionalMetadata":{}},{"type":"org.apache.flink.configuration.IllegalConfigurationException","message":"Overlapping
>>> key prefixes detected (job.autoscaler.target.utilization ->
>>> job.autoscaler.target.utilization.boundary), please replace with Flink v2
>>> compatible, non-deprecated keys.","additionalMetadata":{}}]}{noformat}
We need to make sure that the correct, root error message is reported for
ClusterDeploymentExceptions
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
