This file seems awfully busy… When updating the repo, i usually have to press retry a few hundred times before i get a write lock:
Maybe the resync of the python stuff should only happen on slave startup? Seems like it happens every time a sandbox opens
Yeah, we might have to tone it down. It used to be part of the auto-update flow, might just have to move it back there.
Or if the update was throttled to only check once an hour? Does it recopy every time, or does it just do a file stat comparison to see if it needs to be copied?
I’d have to look at it, I believe it should be checking meta data first. That might not be working properly, however, we’ll investigate.