Learn more about these different git repos.
Other Git URLs
Pagure is usually quick to process various asynchronous tasks and thus people don't have to wait for too long when they want to, for example, fork a repository. However, if the repository is bigger (400 MiB - https://pagure.io/archlinux-packages), it is expected that the forking will take longer. And indeed, it does.
Currently Pagure automatically shows its "sorry for inconvenience" message ("This is taking longer than usual... Sorry for that.") after just couple of seconds of processing.
This is not bad per se, but it could leave people wondering whether there are some infra issues that should be reported somewhere.
Forking of larger repositories will always be slower and I am wondering if there is a way for Pagure how to be a bit smarter about those "sorry for inconvenience" messages. I think fork tasks go to "slow celery queue" by default so maybe have a different threshold for such "slow" tasks and different threshold for normal, quick tasks?
Metadata Update from @jlanda:
- Issue tagged with: RFE
Forking is actually using the fast queue as is everything that is likely to impact the user's experience using the web UI.
The issue is simply that the task system doesn't provide any information as to the state of the task. We're still using redis and it basically only provides: pending/running/done so that's all the information we can return the user.
We could tweak when the message about the task being slow appear but the heuristic to determine how long a task will take before doing is likely not going to be a simple one.
Metadata Update from @pingou:
- Issue tagged with: wishful
to comment on this ticket.