Forums

Hi there! I've come back to my side project (webapp) after a few months. Yesterday I pulled the changes and tried to run the usual python manage.py migrate , npm i && npm run build and python manage.py collectstatic. The npm related stuff seems to run smoothly, but anything python based wont work. After hitting the enter-btn after such a command, the curser just jumps to the next line but nothing is happening in hours.

I'm not really able to locate the issue. It works on my local machine and I can't find anything in the logs except the following: 2024-11-06 00:15:42 spawned 2 offload threads for uWSGI worker 2 2024-11-06 00:16:05 uwsgi_master_manage_emperor()/read(): Bad file descriptor [core/emperor.c line 2442] 2024-11-06 00:16:05 lost connection with my emperor !!! 2024-11-06 00:16:05 unlink(): No such file or directory [core/uwsgi.c line 1662]

I disabled, enabled and restarted my webapp several times. I also killed the existing terminals, but nothing has helped so far.

I did see some topics here where people also have some problems - does Pythonanywhere have some problems atm, or did I somehow blew up my app?

Thanks!

Hi @Nifu, it looks like you were caught up in a outtage on a particular server which hosts your files. Unfortunately it seems like there's a gap in our alerting and nobody was woken up to deal with the issue so it carried on for some time before fixing itself. It looks like it's functioning correctly now but as I've just been notified , we don't currently have any more information about what went wrong. I'll let you know when I know more

Thank's a lot for your quick response!

It looks like it's functioning correctly now

That means my scripts should work again? Because I tried again to run python manage.py migrate but the outcome is still the same. No output or anything else. I did reload the webapp of course.

I'll wait for your update.

Thanks again!

Yeah it should have been working again since 11:30 UTC.

Here's the update I've been sharing around about the problem: We recently experienced an outage on one of our servers, which affected some of your files. Unfortunately, due to a gap in our alerting system, we weren’t notified in time, so the issue persisted until it resolved on its own.

This outage specifically impacted SQLite databases, as they require exclusive access and a lock to function correctly. When multiple processes attempt to access an SQLite database, only one can do so at a time, meaning a lock must be in place to prevent data issues. During the outage, this locking process was disrupted.

We've now addressed the root cause and improved our alerting system to ensure prompt action if anything similar happens in the future.

We apologize for the disruption and thank you for your patience as we continue to improve our service.