Some of the nodes are getting this health check warning: We found one or more duplicate user accounts in the connected user directories.
Platform Notice: Data Center Only - This article only applies to Atlassian products on the Data Center platform.
Note that this KB was created for the Data Center version of the product. Data Center KBs for non-Data-Center-specific features may also work for Server versions of the product, however they have not been tested. Support for Server* products ended on February 15th 2024. If you are running a Server product, you can visit the Atlassian Server end of support announcement to review your migration options.
*Except Fisheye and Crucible
Summary
Some of the nodes are showing the Health Check warning "We found one or more duplicate user accounts in the connected user directories".
Environment
8.19.0 and higher
Diagnosis
When running the curl commands below to check which user is duplicated, nothing is returned:
1
2
3
4
5
6
7
curl -X GET "http://<jira base url>/rest/api/2/user/duplicated/count" -u "adminuser"
Enter host password for user 'adminuser':
{"count":0}%
curl -X GET "http://<jira base url>/rest/api/2/user/duplicated/list" -u "adminuser"
Enter host password for user 'adminuser':
{}%Cause
The Duplicate user accounts health check runs once a week to scan all user directories connected to Jira for duplicate user accounts. An account is considered a duplicate if another active account with the same username exists in another user directory, or if a user has multiple accounts but their only active account is not in the directory with the highest priority. Reference: Health check: Duplicate user accounts
The REST API querying and health check status are working separately:
REST API requests have their cache holding information about the number of duplicate users and their detailed list valid for 10 minutes.
Health check status stores its cache record only about the number of duplicated users, and it is valid for 7 days. So every time we check the health check status, we will see the same information until the mentioned 7 days have passed or until we restart the node.
Nodes are not replicating changes to this cache between them
Solution
The Health check status is per node and remains valid for 7 days. In order to update the results of the health check status, we can:
Restart the affected nodes;
or you can ignore the message, knowing that the scheduler task will clean up the cache on its next execution.
Was this helpful?