We want to hopefully increase our Jira instance's performance by using the Custom Fields Optimization tool, including deleting all our unused fields. However, I've seen the caveat about how Jira won't see a a custom field being used if it's in an archived issue. (https://support.atlassian.com/jira/kb/custom-field-optimizer-and-archived-issues/) We do have a number of archived projects.
Hi @John Schroeder
Welcome to the community!
I would be cautious here.
Atlassian’s own KB says the optimizer does not count archived issues when determining whether a custom field is “used”, so yes, a field can look unused even though it still exists in archived projects/issues.
Because of that, I would treat the optimizer as a strong hint, not absolute proof, when archived projects exist.
My practical approach would be:1. do not delete immediately based only on optimizer output if archived projects matter to you2. prefer “Change context” first where possible, because it is safer than deletion3. export/document the field configuration and contexts before deleting anything4. keep a record of custom field IDs/names so you can trace what was changed later
So yes, if archived projects are important for audit/history, the safer route is to narrow context first and be conservative about permanent deletion.
Hello and Welcome @John Schroeder
I would be very conservative with deletion. Atlassian’s Data Center docs say that when you delete a custom field, it is gone forever including all stored data. They also say that if you delete a context and the field does not appear in any other context, you can permanently lose all field values and that deleted data cannot be restored.
Actually Field Optimizer are little Buggy
[JRASERVER-77748] Custom field optimizer doesn't show projects with archived issues after running a scan - Create and track feature requests for Atlassian products.
[JRASERVER-73207] Jira Optimizer does not consider Archived Issues when Checking if a Field is Used - Create and track feature requests for Atlassian products.
Before you do that, be sure:
Thanks for the advice!
It looks like you're new here. Sign in or register to get started.