reindexing

Giorgi Tsitsvidze

hello everyone

i have some question about Re-indexing , 

before we add in database server  -  +2 CPU in server -  issue Re-indexing took more than 2 hours after add CPU it need 1,5 hour . 

and + when we have Re-indexing process jira is slowing down

it is normal for jira or we need some Improvement ? 

how we can archive in database issue "which is old then let's say 2 years" 

or we must delete issues from jira and had database backup ( or export in CSV file ) and if we need some logs for see this issues late ? 

database : 
6 CPU
16 RAM
issues  : 130 K 
Custom fields : 100
projects : 150 

thanks !   

 

Rachel Wright

Hi @Giorgi Tsitsvidze , I can only respond to part of your question, so hopefully others out in the community can respond too.

I can share that 130k issues and 100 custom fields is a very small number!  I once had a server instance with over a million issues, many hundreds of custom fields, hundereds of workflows, and thousands of users and Jira handled it just fine. (Not that many hundereds of workflows is ever a good idea!)  I can't remember exactly how long reindexes took, but I'm thinking it was hours for a "background" reindex (when the application stays online), and much less for a "lock and rebuild" reindex (where the application is unavailable to users.)  So the length you reported doesn't seem abnormally long to me.  Which type of reindex are you doing?  Can you do a "lock and rebuild" reindex after hours so users are not impacted while the reindex runs?

Have a look at the Jira Server sizing guide to get an idea of how "big" your application is compared to others.  https://confluence.atlassian.com/jirakb/jira-server-sizing-guide-975033809.html 

For your archive question:  The only applications that have a true archival features are Jira Data Center, Jira Cloud Premium, and Jira Cloud enterprise.  Since you have Jira Server, your only option is to export old data, migrate it to a secondary application (which might create more work for you if you need to migrate all data after Server's "end of support" date in 2024), or move the data to a non-Jira long term storage location.  But before you do all this, what are you trying to achieve?  It's true that removing issues will speed up the index, but I'm not convinced (yet) that you need to speed it up.  Are you having any performance issues in the application itself (not related to the reindex time)?

Let's see what others in the community have to say on this.  I'd love to hear from some Sys Admins!

Cheers,

Rachel Wright
Author, Jira Strategy Admin Workbook