Duplicate issue numbers found

[Rob Bradley]

We have duplicate issue numbers generated for new service desk issues. Can issues have new numbers assigned to them without deleting all the issues in a project?

[Rob Bradley · Accepted solution]

We wound up having to create an XML backup of Jira, spun up a new application and database server, and imported from the XML backup, then copied over attachments and avatars

[Nic Brough -Adaptavist-]

How are you generating these duplicates?  What do they look like, and how do you get to them? 

[Rob Bradley]

These are issues generated from the service desk customer portal by service desk customers

We have email and Slack alerts, and I'm noticing some issues have the same key, and so I can't open the first issue generated with that key when I'm in the project associated with that service desk

[Nic Brough -Adaptavist-]

Problem is, Jira physically can't create duplicates, the database is structured to prevent it, as well as the code.

So we need to track down how they are being created, how they are identified and how they appear in your system

[Rob Bradley]

I did an export of the issues in that service desk to a CSV, and it looks like issues with duplicate issue keys have different issue IDs

[Nic Brough -Adaptavist-]

Ok, but what do you see in the UI?  Where are the duplicates there?

 A sample of the CSV, with column headers, would be interesting to compare (drop summary or text fields if you want, we don't need to see them)

[Rob Bradley]

I didn't see them in the UI, but I did see them when exporting to CSV and also saw duplicate keys in the Slack and email alerts

I seems to have stopped after a re-index, but I'm still monitoring the issue

Exporting to CSV now doesn't show duplicate keys