[FLINK-26261] Reconciliation should try to start job when not already started or move to permanent error

• Type: Sub-task
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Fix Version/s: kubernetes-operator-0.1.0
• Component/s: Kubernetes Operator
• Labels: pull-request-available
• Assignee: Thomas Weise
• Reporter: Thomas Weise
• Created: 21/Feb/22 00:05
• Updated: 27/Mar/22 13:16
• Resolved: 09/Mar/22 11:41

Description

When job submission fails, the operator currently keeps trying to find the job status. In the case I'm looking at the cluster wasn't created because the image could not be resolved. We either need the logic to re-attempt job submission or flag the submission as failed so that JobStatusObserver does not attempt to check again. We should also capture the submission error as event on the CR.

Issue Links

• links to: GitHub Pull Request #21
• links to: GitHub Pull Request #46