Migrating Confluence off of Jira's server, what is the best way to go about this?

[Daniel Pasacrita]

Hey everyone, so we're in the process of planning to move our Confluence installation off of our JIRA server, and we were wondering the best way to go about this. I've seen atlassian's documentation on this issue, but I didn't feel the documentation was especially clear or helpful, and seemed to have some errors even.

Basically what we want to do is just move the Confluence installation onto another server, have it point to the same database, and use Jira's user database, just as it does already, with all of the same data and spaces.

Now Atlassian recommends doing stuff like copying installation directories and data directories over, but what I'm wondering is why not just have a fresh install on a new server and then import the backup file into confluence using the built in restore function? Wouldn't that be much easier? Any help we can have on this would be much appreciated. Thanks.

[Nic Brough -Adaptavist- · Accepted solution]

It might be easier, it might not.

Simple fact is you can pick up an entire installation and put it on another server.  Stop it all, copy the home and installation directories to the new server (identical paths, although changing them later is quite easy), and start it on the new server.  As long as your network routes to the databases behind JIRA and Confluence are open to the new server, you are fine.

Or create a new installation and point it at the old database and re-index it after starting.  Or do the export/import thing.

The difficulty here is we don't know how complex your installation of Confluence is, which could determine the best path for you.  (I'm really not sure that Atlassian's docs are unclear, unhelpful or erroneous on this - I've given them to loads of people who have successfully used them to move servers around, so you might want to explain what you think is wrong with them)