if null pointer error in job handler ,io.zeebe.client.impl.subscription.JobRunnableFactory#executeJob will get null as fail command message, so workflow instance will pending at that job without incident.
but if we wrap null pointer exception by RuntimeException, than that job will fail with incident after exhausting retry times.
so i guess null error message will cause something wrong, please confirm it
Hi @aximo, did you test this? My expectation is that a failure with no error message causes a retry on the broker, and after exhaustion of retry attempts, an incident is raised. If it is not, that would be a bug.