The database collation 'latin1_swedish_ci' and table collation 'latin1_swedish_ci' are not supported by JIRA.

kumar June 8, 2016

Dear Atlassian Team,

I have encountered this issue under "Instance Health" plugin. I have googled and everyone says this is a bug in the "Instance Health" plugin. I want to confirm  whether it is a bug or I should fix this issue. Could you please help regarding this issue.

Thanks.

 

 

1 answer

1 accepted

0 votes
Answer accepted
Petar Petrov (Appfire)
Rising Star
Rising Star
Rising Stars are recognized for providing high-quality answers to other users. Rising Stars receive a certificate of achievement and are on the path to becoming Community Leaders.
June 8, 2016

The supported collations for the different databases can be found in this Knowledge Base article. It also describes the impact of using unsupported collation. Using an unsupported collation may work fine, but there are no guarantees.

Nic Brough -Adaptavist-
Community Leader
Community Leader
Community Leaders are connectors, ambassadors, and mentors. On the online community, they serve as thought leaders, product experts, and moderators.
June 8, 2016

I think your google searches have failed you.

It's not a bug in the slightest, the health check is picking up an unsupported configuration that you're using.  I suspect there's a possible bug in JIRA in that it doesn't stop an install or startup when using an unsupported collation, but it's not a bug in the healthcheck.  It's working correctly.

Petar's link explains all the stuff you need to understand why it's reporting this issue to you.

In my experience, it's caused two problems for me.  There are probably others.

  1. XML exported from systems with the wrong collation sometimes cannot be imported into other systems (whether they're using another unsupported collation or a supported one).  The XML can be cleaned and re-imported, but it's better to not to impose those extra steps
  2. When using a user external directory on a multinational site, people in one region were creating new users with perfectly valid names that contained accented characters that English doesn't use.  When they tried to log in, it would not recognise them because the collation could not cope with those characters and changed them to something it could store and use, but, of course, the users had no way of knowing what to type instead of their non-English letters.

    Both fixed by swapping to a valid collation.
kumar June 13, 2016

Thanks for your responses.

Suggest an answer

Log in or Sign up to answer