[SOLVED] Jupyterhub (TLJH): Failure setting up Development Environment: "Failed to start jupyterhub.service."

Jupyterhub (TLJH): Failure setting up Development Environment: "Failed to start jupyterhub.service."

since last week, I am unable to set up a TLJH Development Environment as described here: https://tljh.jupyter.org/en/latest/contributing/dev-setup.html.

Up until January 23 2023 the process went smoothly, but after rebuilding the image last Friday (27 January), after running bootstrap.py in the container, the server doesn't spawn. Browsing to my localhost:12000 I just see the message "404 page not found".

I am running TLJH in a container in an Ubuntu 22.04 host, running Docker version 20.10.21. On the same host, when using the image I build before January 23 with the same method I can spawn the server ok. There were no new commits on the TLJH repo since, so the issue should not be there.

Here is the output of journalctl -u jupyterhub -f ran inside the container:

-- Logs begin at Mon 2023-01-30 12:44:34 UTC. --
Jan 30 12:44:41 87f857896c91 python3[64]:         raise exc.ArgumentError(msg, code=code) from err
Jan 30 12:44:41 87f857896c91 python3[64]:     sqlalchemy.exc.ArgumentError: Column expression, FROM clause, or other columns clause element expected, got [1]. Did you mean to say select(1)?
Jan 30 12:44:41 87f857896c91 python3[64]:     
Jan 30 12:44:41 87f857896c91 systemd[1]: jupyterhub.service: Main process exited, code=exited, status=1/FAILURE
Jan 30 12:44:41 87f857896c91 systemd[1]: jupyterhub.service: Failed with result 'exit-code'.
Jan 30 12:44:41 87f857896c91 systemd[1]: jupyterhub.service: Scheduled restart job, restart counter is at 5.
Jan 30 12:44:41 87f857896c91 systemd[1]: Stopped jupyterhub.service.
Jan 30 12:44:41 87f857896c91 systemd[1]: jupyterhub.service: Start request repeated too quickly.
Jan 30 12:44:41 87f857896c91 systemd[1]: jupyterhub.service: Failed with result 'exit-code'.
Jan 30 12:44:41 87f857896c91 systemd[1]: Failed to start jupyterhub.service.

Since the logs show an error in sqlalchemy, I tried dowongrading sqlalchemy from 2.0.0 to 1.4.46, and then reloaded the hub, but that didn't solve the problem

Is anyone else experiancing the same issue? Any ideas for further troubleshooting?

Thanks!

EDIT: I applied the solution described here and it worked.

Commands to run: pip3 install --upgrade 'SQLAlchemy<2.0.0' tljh-config reload hub

Solution

I applied the solution described here and it worked.

The reason seems to be an issue introduced in SQLAlchemy 2.0.0, is resolved by downgrading SQLAlchemy.

Commands to run: pip3 install --upgrade 'SQLAlchemy<2.0.0' tljh-config reload hub