Howto keep content ownership when switching user directories?

[Heinz Juergen Letsch]

Hi all,

one of our customers (running Jira/Confluence/Crowd DataCenter) has to switch LDAP directories due to some corporate changes. At the same time the customer wants to get rid of Crowd, ie. the current Crowd directory used in Jira and Confluence is getting replaced by a single Azure AD directory directly connected using a 3rd party SSO App. Usernames will also change from "custom" to email adress in this process.

My question now is the best/easiest way to keep ownerships of both Jira and Confluence content in this scenario?

I found this article which sounds quiete reasonable. However I found the statement:

hese steps are intended to fix Jira content ownership after an environmental change to LDAP, wherein the original directory is no longer used and the new directory will be used going forward. This only applies when migrating from one external directory to another, where the usernames change and both directories have already been added to Jira (meaning users have created content under both usernames). 

And now I'm wondering whether there's another (hopefully simpler) approach for the case that the users created content only under one username? 

I already tried another approach by mapping the users to the new directory on  DB level (keeping the original user_id values). This basically worked as expected, but unfortunately prevents the AAD sync to work correctly...

Anyone has experience with this kind of problem? Any help would be greatly appreciated!

Thanks
Jürgen

[Pramodh M]

Hi @Heinz Juergen Letsch 

Have you got the resolution for the above requirement?

Here's one we have tried some time back

https://confluence.atlassian.com/jirakb/move-local-group-memberships-between-directories-in-jira-server-289277867.html

Thanks,
Pramodh

[Heinz Juergen Letsch]

Hi @Pramodh M ,

thanks for your followup and the link. 

In the meantime I found a solution on my own. Here's a brief outline what I did.

• prepare a mapping file (CSV) of old usernames (Crowd) and new usernames (AD)
• deactivate AD user directory
• move all Crowd users to internal directory (stop instance; just change the directory_id using SQL). This will allow renaming users in the next step.
• rename users using a groovy script (REST calls on restarted instance) and the above mentioned user name mapping. 
• activate AD directory
• that's basically everything

As we do not touch the actual user id (and especially the key!) all content is still bound to this one user which now can login with the AD directory activated.

Best
Jürgen