Re-indexing and Lucene Health best practices

[Matt Strandberg]

Hello,

We've had several occurrences of Lucene health check failures over the past couple of weeks.  Sometimes a reindex fixes the issue, sometimes it doesn't and often the index issue ends up turning into a larger system issue that impacts our users.

Are there any recommended steps to follow to begin troubleshooting deeper issues with the index?  Are there any best practices anyone can share regarding regular maintenance, monitoring, etc.. of Lucene?

Instance info:

• Jira Software Version 7.5.0
• Server license, hosted in AWS
• ~3k users
• 316k issues
• ~200 custom fields
• ~350 projects

[Gonchik Tsymzhitov]

Hi! 

I think you need to find the app or automation which corrupt the index. 

And check latency as well