[SLING-4635] Topology related jobs shouldn't be scheduled if topology is unclear/not available

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: Commons Scheduler 2.4.6
• Fix Version/s: Commons Scheduler 2.4.8
• Component/s: Commons
• Assignee: Carsten Ziegeler
• Reporter: Carsten Ziegeler
• Created: 17/Apr/15 13:48
• Updated: 13/Jun/15 16:12
• Resolved: 20/Apr/15 12:36

Description

If a job is scheduled to be run only on the leader or only on a single instance, it is currently scheduled even if the topology is not available or changing.
This is wrong as this can lead to the job being scheduled more than once in a cluster in such situations.

Issue Links

• is blocked by: SLING-4638 isCurrent() not always correctly set