Primary lock is held by another instance is logged on passive nodes in Bamboo DC
Platform notice: Server and Data Center only. This article only applies to Atlassian products on the Server and Data Center platforms.
Support for Server* products ended on February 15th 2024. If you are running a Server product, you can visit the Atlassian Server end of support announcement to review your migration options.
*Except Fisheye and Crucible
Summary
The below log line is printed in the $BAMBOO_HOME/atlassian-bamboo.log on a secondary passive node in a Bamboo DC setup and may continue even after the primary has been shutdown:
2021-08-01 12:01:00,246 INFO [localhost-startStop-1] [ClusterLockBootstrapServiceImpl] Primary lock is held by another instance, suspending....
2021-08-01 12:02:00,821 INFO [localhost-startStop-1] [ClusterLockBootstrapServiceImpl] Primary lock is held by another instance, suspending....
2021-08-01 12:03:00,819 INFO [localhost-startStop-1] [ClusterLockBootstrapServiceImpl] Primary lock is held by another instance, suspending....Solution
This is a normal part of Bamboo DC (8.0)'s active-passive locking configuration. The active node will claim a database lock for 5 minutes and continuously updates this lock while it is online. Even after the primary goes down or is shutdown, the lock will remain in place for 5 minutes to give the primary a chance to reclaim it (e.g. in the event of a restart). Once 5 minutes has elapsed, the lock is available to be claimed by another node.
At this point there is no way to tune the timing or cluster lock mechanism, however we do have an open feature request for that below:
- BAM-21400 - Getting issue details... STATUS