should attempt to mirror failing branches less and less frequently

[Michael Hudson-Doyle]

This is talked about in other bug reports, but there's not one specifically about it:

When the supermirror-pull.py mirror script fails to mirror a branch, it should increase the delay before the branch is tried again (probably exponentially), to reduce run times and noise in the logs.

Unfortunately, it involves a scary db query in the authserver...

This is related to bugs #74033, #74031 and #127706 at least.

[Joey Stanford (joey)]

+1 along with some sort of cut-off where we stop mirroring altogether and force the user to reactive the branch via Launchpad.

[Jonathan Lange (jml)]

I'm a little unclear on the exact state of the bug.

There's a branch that implements the backoff, but I think it depends on certain user interface fixes that I haven't done / defined clearly yet :( I'd like to try to get this into 1.1.9.

[Jonathan Lange (jml)]

So this is fixed, except for the part where users can re-activate the branch via Launchpad.

Branch for that is going in for a late review.

[Jonathan Lange (jml)]

This has been released.