Incorrect Collation for JIRA MSSQL db. What are the risks?

[Eric Bendixen]

I recently setup a test server to update our JIRA installation from an outdated version and bring it up to version 7.11.2.  As part of this test upgrade, I asked our DBA to restore a copy of our production MSSQL database into our development environment to be used as a testing db.

Everything with the test upgrade went smoothly and I am now working in 7.11.2, but I am now getting alerts that our database collation is set incorrectly to SQL_Latin1_General_CP1_C1_AS.  I have to assume that this is the collation that has been used for the past few years on our production database.  While I would like to get it moved so that we are using the correct collation format, our DBA is out of the office for the foreseeable future. I'm just curious as to what the risks would be in continuing to run using this collation format.  Thanks!

[Mauricio Karas]

Hello, Eric. Thanks for reaching out to the Atlassian Community.

We know that some issues with invalid characters in certain languages could happen and also installation and upgrades could fail due to invalid column names, as mentioned in this KB article.

But the major thing about setting the right collation is that Jira was built and tested on this collation, so even though it could run on a different setting, we can't be so sure that some issues will not occur, therefore the application can't be supported with this setting.

So we strongly recommend you to set your database to the correct collation. You can check the KB article below that has all the steps necessary to change the collation on MSSQL:

• How to fix the collation of a Microsoft SQL Server JIRA database 

Kind regards,
Maurício Karas