[SLING-7756] Improve logging for Job retries

• Type: Improvement
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: Event 4.2.10
• Fix Version/s: Event 4.2.18
• Component/s: Event
• Assignee: Konrad Windszus
• Reporter: Konrad Windszus
• Created: 27/Jun/18 12:55
• Updated: 17/Apr/23 23:19
• Resolved: 11/Jan/19 08:25

Description

Currently a JobConsumer returning JobResult.FAILED is usually retried (https://sling.apache.org/documentation/bundles/apache-sling-eventing-and-job-handling.html). Unfortunately this is not exposed in the logs, so in case a JobConsumer logs an error before returning JobResult.FAILED the job is silently retried without any particular log message. I think each failed job should lead to a WARN in the log (in case the job is retried) with the attempt number and outstanding attempts in the log message. In case a job is not retried, an ERROR should be emitted to the log for every failed job.

Issue Links

• links to: GitHub Pull Request #2