Hello,
The "Clean Journal Entries" scheduled job (formerly known as "Optimise Indexing") is very demanding on system resources according to some fairly ancient documentation. However, if you turn off the job scheduling, the size of the journal entries (which are one of the ways that search indexes for nodes in Data Center stay in sync) can potentially grow indefinitely in size, causing you to run out of disk space.
Unfortunately, there's really no public documentation available except the above on precisely what occurs during this job, and I'm not sure how much of the documentation on "Optimise Indexing" applies to "Clean Journal Entries". However, if you have a large Confluence instance, with 3 or more nodes, I would expect some significant performance degradation while this job is occurring based on my experience and the little information I can find online.
If you go to Scheduled Jobs in the Confluence General Configuration, you can see the time (in milliseconds) that it takes to complete the Clean Journal Entries job. What's the time listed here?
If the time listed doesn't match up with what you're expecting to see (1000000 to 1500000 milliseconds) then this job may not be the source of the performance hit you're seeing - by default, Confluence is scheduled to run a backup at 2 AM as well.
You can try modifying the cron expression for this job to run on a weekend, but you will need to closely monitor disk space, especially if you're operating within tight constraints.
You must be a registered user to add a comment. If you've already registered, sign in. Otherwise, register and sign in.