Sentry <-> Jira assignee sync not kept when merging issues

[MaximeLaurenty]

Environment

SaaS (https://sentry.io/)

Version

No response

Link

https://spinergie.sentry.io/issues/4081093683/activity

DSN

No response

Steps to Reproduce

• setup JIRA integration with assignee sync on
• create 2 issues in sentry
• assign one (Α) to someone
• create a JIRA ticket for the other (B)
• merge the 2 issues in sentry

Expected Result

After merging, the JIRA ticket gets assigned to the assignee in sentry.

Actual Result

The issue in sentry has an assignee but the JIRA ticket has none.

Even if the JIRA ticket for B and the Sentry issue A had different assignees initially, given there are already rules to keep only one assignee and one ticket amongst the different merged issues, a precedence order already exists, hence the sync to JIRA should be triggered in the end. Note that if one un-assigns the issue and re-assigns it after merge, the sync is then triggered, but it's a pain to have to do this.

[getsantry[bot]]

Assigning to @getsentry/support for routing, due by Monday, April 24th at 9:15 am (sfo). ⏲️

[getsantry[bot]]

Routing to @getsentry/ecosystem for triage, due by Tuesday, April 25th at 1:37 pm (sfo). ⏲️

[Dhrumil-Sentry]

Hi @MaximeLaurenty thanks a lot for the feedback, I agree the Jira integration should handle this case and will add this to our backlog, but could you please share more details on why your merge issues?

Do you find Sentry's grouping to be inaccurate often?

[MaximeLaurenty]

• For backend projects the grouping is close to perfect. I have only example in mind: when our error_message includes some user info, eg. "user X should not be able to reach action Y". For our main usecase (rare event) it is clearer than a generic "user should be not be able to reach action" + then having to look through tags to find the user and action. When the issue is related to poor code and not to user though, we can get 1 message per user, which we want to merge then.
• For our frontend project, we have to merge issues quite often because they don't share the exact same stacktrace (often they share their last 10-15 levels but 2-5 upper levels differ). The "similar issues" tab is quite useful then (though we don't merge everything that appear there).