The particularly odd thing is that the TSTP "project" doesn't exist in our instance (or so all the UI views agree), but I also cannot create it because it already exists. Also, the only way it can have gotten into the merged dump file is from the "mirrored" SVN repo I had to create against the OnDemand SVN repo.
This is critically blocking our migration onto the Atlassian platform, and could cause us to reconsider our decision.
OK, there's now a bug in Atlassian's own Jira for this.
If anyone else hits this problem, the solution is to use 'svndumpfilter' to remove the offending orphan project from the SVN dump file you create from the merged repo before uploading it to the WebDAV folder and (trying to) import it to your AOD SVN repo (overwriting the existing contents).
On top of that, the projects in your SVN dump should match the number/keys of the projects you already have on OnDemand. For reference - http://confluence.atlassian.com/display/AOD/Importing+Versioned+Data+into+Subversion
The convention is to have a top-level directory per project, whose name is the same as the project key. Please note, all Subversion content in your dump file must be associated with a OnDemand project. If a project does not yet exist, you need to create it before you import.
Connect with like-minded Atlassian users at free events near you!Find a group
Connect with like-minded Atlassian users at free events near you!
Unfortunately there are no AUG chapters near you at the moment.Start an AUG