New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
updategit does nothing #2400
Comments
The |
It doesn't seem to make a difference either if I use Also, it seems that using the API |
We noticed the same behavior. The
There are some details in celery's log:
Pulling the latest updates via |
Interresting, I don't see those on my end. Just the INFO that the task was received. |
This might help diagnose Celery configuration issues like in #2400. Signed-off-by: Michal Čihař <michal@cihar.com>
@nblock Not sure if your problem is related, that most likely looks like something wrong in your Celery setup - using older protocol might help (see https://stackoverflow.com/a/48648079/225718) or downgrading redis package (see celery/celery#5175). @BhaaLseN The /var/log/celery/weblate-w1.log log is for the dispatcher only, the individual tasks log into /var/log/celery/weblate-w1-*.log (per worker), so better check these logs for errors. |
The log (for DEBUG level) should look like:
|
Uh, ok, didn't know they were different. I did see a permission error in there (and fixed it), but I'm not sure if that was the only thing causing problems. I'm on the road atm and don't have full source access, so I won't be able to verify that until monday or so. I'll leave this open for now if you don't mind; mainly as reminder for myself. |
Alrighty, over night I had another error in the celery worker log:
That WRITELOCK file existed (along with 5 others), it was empty (zero bytes), and the owner was I removed those files, double-checked permissions (which are now owned by a After that, all that was left was an SSH authentication issue (caused by the private key having permissions that didn't allow celery to read them; mostly because they were still owned by With that, updates now work again. Slightly tedious to debug though, since there is so many places where things can go wrong, and no clear "look at that log file" hint. |
This usually indicates some configuration issue, eg. celery running as different user than wsgi. See #2400 Signed-off-by: Michal Čihař <michal@cihar.com>
This is something Weblate could check automatically and I've implemented the check in e1b0c87. Thank you for your report, the issue you have reported has just been fixed.
|
Thanks for your hint. It was a deployment issue. For some reason, I used celeryd from Stretch which is too old and has some issues. Switch to the recommended configuration as per the current documentation and it seems to work fine now. |
Describe the bug
The management command
updategit
doesn't seem to work correctly at this point (no updates happen in the working copy)To Reproduce
Steps to reproduce the behavior:
./manage.py updategit --all
Expected behavior
The working copy should be updated and translations should synchronize to see the changes on the given component page.
Server configuration and status
Additional context
It took a while for us to notice this, since we had a slow month since the update; as well as some environmental power failures that caused Celery to not work.
But even after restoring,
updategit
seems to have no effect.commit_pending
seems to be a little quirky (which could be #2396) as well.Both
updategit
andcommit_pending
run on an hourly cronjob; mostly because we don't have infrastructure in place (yet) to trigger updates on demand. On that note, using the API operationpull
on the repository does update correctly; and we're likely to use that as a workaround in the mean time.In case it matters, this was a Git update from
weblate-3.1.1
toweblate-3.2.2
(to narrow down a range where it could've happened)The text was updated successfully, but these errors were encountered: