Web not connecting to the db on docker

[pjmoura577]

[Xiphoseer]

Are there more logs? In theory this only happens if a previous migration crashed after creating the notes column

[PlatinPython]

I've run into the same error. The following steps are what I did to arrive there:

1. Install Docker Desktop and move the client to the client subfolder
2. Download docker-compose.yml and .env.example (renamed to .env)
3. Set ACCOUNT_MANAGER_SECRET to a "SHA 256-bit Key" and MARIADB_PASSWORD to a passphrase containing only a-z and -
4. Run docker compose up -d in the folder containing client, docker-compose.yml and .env
5. Start the dlu-darkflameserver-1 container again (It stops for some reason)
6. Run docker exec -it dlu-darkflameserver-1 /app/MasterServer -a
7. Try to log in via the Nexus Dashboard (This leads to an internal server error)
8. Try to manually run flask db upgrade in the dlu-darkflameweb-1 container.

Logs: https://gist.github.com/PlatinPython/59a8c9c14d7904c1424cc377a003f789

Running Windows 10 Pro 22H2 and using the WSL 2 backend in docker.

[Xiphoseer]

Hi @PlatinPython thank you very much for the detailed report. Based on the errors in the logs, I think you may be hitting MariaDB/mariadb-docker#331 (aka https://jira.mariadb.org/browse/MDEV-24189, aka microsoft/WSL#8443), which explains why I can't reproduce it on linux. Can you check whether the reproduction in microsoft/WSL#8443 (comment) applies to you?

EDIT: Can you share more details on your docker setup? Are you on windows? Are you using the WSL-2 or the Hyper-V backend for docker / docker-desktop? Can you try the other one?

[PlatinPython]

As far as I can tell, the reproduction linked does not apply to me, as I don't get error on fstat errno 2 anywhere in my output.
I had already added the details of my docker setup to my original comment. I'll give the Hyper-V backend a try and report back.

[PlatinPython]

Docker with the Hyper-V backend seems to work fine. Login works and flask db upgrade also runs without errors.