After cloning a server all application links are offline

Gerald Schneider

I cloned a production server to test the upgrade to newer versions of JIRA, Confluence, FishEye and Crowd before rolling it out on the production environment.

After cloning and changing all references to the new hostname everything seemed to be working, except the application links were dead. Trying to relocate them to the new hostname only resulted in the error message

The application at URL '...' is not responding. Please confirm that you want to use this URL.

Gerald Schneider

After a lot of searching (the logs were not helpful) I found the solution rather by chance in the Apache configuration.

The server is configured to use HTTP basic auth for authentication. The new IP of the server was missing in the "Allow from" directive of the auth configuration.

AuthBasicProvider crowd
 CrowdAppName apache
 CrowdAppPassword XXXXX
 CrowdURL http://localhost:8095/crowd/
 Require valid-user
 Order allow,deny
 Allow from 127.0.0.1 x.x.x.x y.y.y.y

After adding the IP to the Allow from line the application links worked immediatly.