sqlite race condition: database is locked

[Svagtlys]

I set up DAVx5 connected to Rustical via an app password, and it was working great. Now, after I messed around with some WebDAV-Push stuff, I'm getting a 500 Server Error from DAVx5 and the following errors in Rustical

2025-10-07T17:27:43.793710Z ERROR http-request{otel.name=PUT /caldav/principal/MyUser/35477676-3327-4e00-b615-21293a339aec/53c05efd-5bff-4542-b935-ecc72753d5d5.ics ua=DAVx5/4.5.4-ose (dav4jvm; okhttp/5.1.0) Android/16 status=500 Internal Server Error}: rustical::app: server error
2025-10-07T17:27:43.793739Z ERROR http-request{otel.name=PUT /caldav/principal/MyUser/35477676-3327-4e00-b615-21293a339aec/53c05efd-5bff-4542-b935-ecc72753d5d5.ics ua=DAVx5/4.5.4-ose (dav4jvm; okhttp/5.1.0) Android/16 status=500 Internal Server Error}: rustical::app: something went wrong

[lennart-k]

Thanks for reporting,
unfortunately I've also had this once and I couldn't reproduce it since then.
If you by chance have the corresponding logs from DAVx5 that would be amazing.
In v0.9.11 I'm now making sure that internal server errors will also be completely logged to the console and not just to the response body, that would be quite interesting if you know how to reproduce that error 😅

It could either be that the UID in the calendar object and the filename don't match or some weird error in the calendar store.

[Svagtlys]

Well looks like the 500 error has cleared up now, I'll re-open if it happens again and I can grab the DAVx5 logs...

[Svagtlys]

debug-info.txt
davx5-log.txt

The issue recurred for a moment last night, and I think I was able to grab the logs of it happening. It seems intermittent, though, so I wouldn't expect this to be a priority, more just re-opening this to sate my curiosity.

[lennart-k]

Thanks, that's extremely helpful. The error seems to be from sqlite:

error returned from database: (code: 5) database is locked

I'm probably running into a deadlock for some reason, most probably my fault, I'll have a look at why that happens.

[lennart-k]

Or is there a chance you had a sqlite shell with the database open? The sqlite docs seem to suggest that that could've been the problem.

[Svagtlys]

I have not accessed sqlite in the docker container, and I haven't messed with the db file at all. The only thing I can imagine is that something about me sending webdav-push subscription requests may have caused issues? But I haven't send those requests again in a few days, and I'm still today getting the 500 error.

[lennart-k]

Sorry for the frustration, I'm quite certain it's not your fault.
It might be some race condition since internally a connection pool to sqlite is used.
At least it seems to easily be resolved by the client by just trying again. :)
Until start of December I have to focus on my thesis but afterwards I'll fix this as soon as possible :D