Comment 3 for bug 210031

I hope so. The "only branch unlock" thing I landed should go some way to reducing the instances of this bug, but I'm not sure whether it will fully fix it.

Essentially, we made the code only request a mirror on branch unlock. If the client unlocks the branch twice during creation, say, then the bug will still continue.

When I get back on Monday I'll gather all the bugs related to this issue. Not sure how to QA without real-time access to staging OOPS logs.