JIRA custom fields and performance

[DI2E Licensing]

We've heard that multi-line text custom fields adversely affect JIRA performance.

(http://blogs.atlassian.com/2015/05/webinar-recap-five-secrets-jira-performance-scale/)

If we always configure these custom fields to only be applied to a single project will that help?  We have a couple projects where are customers are asking for multiple multi-line text custom fields and we are weighing these requests against our JIRA performance.

Any advice welcome.

Thanks.

[Nic Brough -Adaptavist- · Accepted solution]

Yes, it will help a little bit because JIRA is structured in a way that checks if there is anything to read first.  When the field doesn't even exist for the project (or issue type), it can stop work at that point, rather than asking for the data to be retreived (and having to process either "no data" or "here's some data").  There's also something about the specific way it is stored in indexing which helps the speed when fields are project-specific, but I'm not going to pretend I understood the rest of the conversation I had about that one.

It will be faster, but you may struggle to measure it unless you're on a big, heavily loaded, JIRA.

[DI2E Licensing]

I have a follow on question . . . Are read-only text fields better from a performance perspective than a regular text field? Or is it so negligible??? Thanks for you help.

[Nic Brough -Adaptavist-]

I don't think the answer to that matters, as read-only fields are different to text fields - there are cases when you need read-only fields and others when you need ones the users can actually use.