-
Bug
-
Resolution: Timed out
-
Low
-
None
-
5.9.1
-
1
-
Severity 3 - Minor
-
Steps to reproduce:
- Set up a one node cluster
- Either populate the instance with a lot of data or disable index flushing job so that reindexing will block index flush from happening during startup
- Stop the node
- Remove ${CONFLUENCE_HOME}/journal/main_index file
- Start the node and let reindexing complete
- Restart the node
Outcome: Reindexing is triggered again
Suggested fix: Update journal id to be the id of the latest entry in journalentry table at the beginning of reindexing.
- relates to
-
CONFSERVER-40383 Index recovery can try to perform reindex too early
- Gathering Impact