Reindexation of two machine

[Diego Martín González]

Hi,
Reindexation of two machines with jira 8.2.3 takes 23 hours, this is not operational, we need to reduce that time considerably because we support 7x24 and we cannot lose service for 23 hours.
We have thought about reinstalling Jira again on the servers, once we put two new nodes, but there would have to be a not so radical solution to improve the reindexing.
There is a way to improve reindexation times?
Thank you,
A greeting

[JimmyVanAU · Accepted solution]

Hi Diego,

Welcome to the community. 23 hours is a long time for reindexing. Can you tell me a bit more about your environment?

You mention 2 servers - are you running Jira Data Center? Can you confirm this is the case, or are you running two separate instances of Jira server? If you're using Jira Data Center, you should be able to have zero down time while reindexing if one server can handle all the traffic. Otherwise we may need to look at scaling horizontally (add more servers).

Can you tell me a little more about the size of your instance? How many users, projects, issues and custom fields do you have? Are there a few large projects and many small ones? Or are there lots of small projects? With Jira 8 (Data Center), archiving should definitely be a consideration

Most importantly, you should also look at why you're needing to reindex constantly. Are many many custom fields being added? Do you have a lot of marketplace apps?

If you let us know a bit more about how things look, then we can provide some suggestions.

Cheers, Jimmy