How Jira define a unique user and identify a unique user

[sc]

Dear all,

We are using Jira Data-Center and we have issue with unique users in Jira database.

It seems to me that Jira identify unique users by its user_key and lower_user_name, is it really the case ?

Because sounds to be by doing so we have duplicate user account in database for same user

Thanks for clarification

regards

[fran garcia gomera]

You can define the order of the user directories. I think only the first one would be used in case of a repetition.

Yoou could check if that is happening with a query to cwd_user to get possible duplicates in low_user_name.

[sc]

My AD directory is define first in the order list from the UI but that does not means that records in cwd_user are in same order..

What type of querry are you thinking about ?

[fran garcia gomera]

cwd_user has a field called directory_id, i think jira gives precedence to rows ordering by that directory, I would say even in loging into the app.

something like this

SELECT lower_user_name, directory_id, COUNT(lower_user_name) 
FROM cwd_user
GROUP BY lower_user_name
HAVING COUNT(lower_user_name) > 1

[fran garcia gomera]

In relations between tables jira uses lower_user_name as id of the user (for instance in worklog table, or as reporter, creator or assignee in jiraissue table or in changeitem via changegroup or even cwd_membership) idon't remember any relation defined using the user id.

[sc]

This is wierd if it is using the field lower_user_name , because you could potentially have different lower user name field for a same user internal and external for instance, and if internal is scan first it will take this entry instead of the syn AD.

Anyway to avoid this ?