-
Type:
Bug
-
Resolution: Not a bug
-
Priority:
Low
-
None
-
Affects Version/s: 7.11.2
-
Component/s: User - Management
-
None
-
1
-
Severity 2 - Major
Issue Summary
When trying to do an incremental sync against a Load Balanced Active Directory LDAP on embedded Crowd 4.0 or greater invocation id is checked and a full sync is forced if the AD node changes.
Steps to Reproduce
- Create an AD connector to a Load Balanced AD instance
- Run incremental syncs and wait for a node
Expected Results
The incremental sync runs regardless of which AD node is being used.
Actual Results
The below warning is thrown in the atlassian-confluence.log file and a full sync is initiated:
2021-04-30 01:45:43,142 INFO [Caesium-1-2] [directory.synchronisation.cache.UsnChangedCacheRefresher] synchroniseChanges Last incremental synchronization took place for AD instance with invocation id 'a013b7a9-da82-4221-9f3e-cc3dda107796', current instance has invocation id of '8b71bd78-d189-41b3-b9d1-d4819a4cd115', falling back to full
Workaround
To work around this an AD node must be accessed directly.