Deleting the Journal IDs on Nodes will cause Indexing to run on restart
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
Problem
If you clear the Journal folder on restart in your Confluence Data Center instance, then the system will perform a full re-index on every restart. This is not desirable in large enterprise installations where the Content Indexing process may take many hours to complete.
Diagnosis
After every restart of a node, Confluence will immediately being reindexing content within the application. This is contrary to expectations, as the Data Center nodes should not be reindexing on restart unless the instance was performing a full rebuild of the indexes from scratch.
Cause
If the edge_index.id and main_index.id are deleted from the journal folder in the Confluence Home directory, then the two nodes will not share the same journal ID for the indexes. If the IDs do not match between the nodes, then the instance will begin a complete content reindex to restore the parity between the nodes.
Resolution
To resolve this issue, do not remove the contents of the journal directory in your instance.