Performance impact using labels

[Hans Meijer]

Hello,

I am looking for some insight on how JIRA is performing when using labels. Who has experience with performance issues using labels?

Would we have a performance issue if we have 20,000 labels considering that we have about 200k tickets, and 200 projects in a single instance?

What other information is relevant to make a statement?

Thanks in advance, and kind regards,

Hans

[Matt Doar _Adaptavist_ · Accepted solution]

It's probably going to work, but test it first. I worked on a DC instance with 10M issues, 800 projects and tens of thousands of labels, and saw no performance problems due to labels. Best to have the labels set by scripts though, since people make typos

[Callum Carlile _Automation Consultants_ · Accepted solution]

Hi @Hans Meijer ,

Some info on this can be found here. The answer is "it depends", and as it says in the link - "There is no fixed number that would be considered as large amount of labels, as it would be dependent on each Jira environment's processing power and capacity."

[Hans Meijer]

Thanks Callum for the link. I understand the "it depends" remark.

I am hoping to hear from community whether 20k labels is a lot, and whether this made them decide to disable the feature, or found the work-around useful.

Thanks and regards,

Hans

[Radek Dostál]

200k issues and 200 projects would be small compared to big companies that span across thousands of projects and millions of issues.

"it depends" is as is, depends

Not sure if it's default but labels are stored in 'label' table which I see here has indexes for all three, the issue id, the label itself, and field id. So in this case, there wouldn't be a problem really even if it were in millions.

The size of that table is 13.5 mil.

Never had a problem with labels.

Then again this specific instance is huge and runs on physical HW. So, it depends, but 20k is nothing in terms of SQL, especially when there are indexes.