Migrating Jira/Confluence new server with Crowd cannot login

Alex Moozhayil

Help! Can someone tell me what I am doing wrong. We are moving to AWS, I install Jira/Confluence/Crowd on AWS. After I completed my install, backup(old server) and restore(AWS) my jira instances. I login into CROWD and setup the Jira configuration, however I am unable to login. 

I have tried going into the SQL DB, and changing the users table to use the internal jira users versus Crowd. Even after this I am unable to login. So my question is, what is the best way to migrate to a new server with Crowd, Jira, and Confluence? 

 

Andy Heinzer

What tends to happen in cases like this is that your old setup has Jira connected to a specific IP/Address for Crowd in order to handle all the authentication. And since that address is likely changing due to this migration, the settings that handle this in Jira need to be updated to let these users login.

For this reason, I would recommend that you leave the Jira internal user directory enabled, AND have a Jira system admin level account in that user directory that is uniquely named (ie a username that does not exist in Crowd, LDAP, or other user directories).  By having this kind of configuration, you can at least then ensure that you have at least 1 admin that can login even if Crowd fails or is unreachable for any reason.

One way through this would be make this change on your old system now, create the backups again, and restore these changes to the new platform.

However the more direct solution at this point would be to follow the steps in Retrieving the Jira Administrator.  This document explains how you can find all the group(s) that have this Jira system admin permission.   In your case, I would look for a user that exists in the Jira internal directory (has a directory_id value of 1).  If you can't find an internal Jira administrator, the guide also has steps on how you can insert such a user account via SQL directly.