what actually happens when you connect two jira instances to one database?

[Elifcan Cakmak]

Hello community,

One of our customers accidentally connected a new Jira instance to the old used production Jira instance's database. We immediately shut down new Jira when we realized this. However for a period, two Jira instances ran connected to the same database. What are the actual results for this action?

Problem: Service desk project stops creating issues that are created via email after this. But even though the issues are not created, reporters are still getting email saying that their issue is created. Also these emails include old issue keys that are already opened before. We think this is because of the mistake above but I cannot find what the actual results of connecting two Jira instances to the same database. EDIT: turns out this problem was because the other Jira was using the same e-mail that is used to create issues via mail.

BUT STILL, There is no useful information in the documentations about what would happen in the case above. We don't know what we might run into in a few days caused by this incident, so we want to be prepared for any issues. 

Any ideas are appreciated.

Thanks.

Elifcan

[Jorden Van Bogaert · Accepted solution]

Hi @Elifcan Cakmak 

I assume that both Jira's may have had the possibility to overwrite each others data. I also believe that this may cause the auto-increasing ID for objects in the database to eventually run ahead of one of the two jira instances, making Jira unable to create new issues (as it will try to create an unique ID that already exists). Not talking about the issue key here, but the actual ID that is stored for objects in the database.

I think worst scenario will be lost data or corrupted data (different custom fields across two instances, with their own data type, let's say free text vs date picker, may result in free text being forced into a date entry. But I believe for this, the database would return an error on the moment it's actually trying t do it.

Jira may also be unable to start if the database stored 'jira version' is not the same as the actual jira version starting up.

These are just my assumptions, I'm not sure how well Jira prevents itself from making such mistakes in the database, but I believe most exceptions are thrown on runtime and if you don't experience any trouble now, I would think the worst you're going to see now is lost or overwritten data (including overwritten filters, board configurations, etc.)

Kind regards
Jorden

[Elifcan Cakmak]

Hey @Jorden Van Bogaert  

Thank you very much for the detailed explanation. These are my assumptions too, but I wanted to get confirmation.

Best,

Elifcan