With containerd.io 2.3.3-1 on Debian 13: Getting error "Couldn't move process x to requested cgroup '/system.slice/docker-y': Device or resource busy

[KomtGoed]

Hello,

On v1.4.0, I am getting the below errors in syslog since yesterday.
Unifi OS as such seems to remain running fine as far as I can tell when logging into the web interface.

Not sure whether this a unifi-os issue or containerd upstream...

Upgraded my Debian Trixie host to 6.12.101+deb13-amd64 yesterday.
Upgraded Docker to 29.7.2 and containerd.io to 2.3.3-1 at the same time.

Already before the reboot with only new Docker and containerd.io running, I started getting this kind of error messages
Aug 8 09:47:31 HOST systemd[1]: docker-e82207515b33dac96abb17591047fc1a49fbd8ef8dfe989cf919e72e61cb41c8.scope: Couldn't move process 1492530 to requested cgroup '/system.slice/docker-e82207515b33dac96abb17591047fc1a49fbd8ef8dfe989cf919e72e61cb41c8.scope/': Device or resource busy

Rebooted the server, no help.

Did a hard recreation of the unifi-os container (shutdown, delete image, start again).
Besides a new container ID in the logs, nothing changed.

Using cgroups in another container as well (e.g. cpuset), no errors.

Downgraded Docker to 29.7.1, force-recreat'ed the container, no help.

Downgraded containerd.io to 2.2.6-1... error disappeared.

[lemker]

Is containerd 2.3.3 in the Trixie package repo or did you install it manually?

[KomtGoed]

Good question, in two ways:
I indeed use "https://download.docker.com/linux/debian trixie stable" as the repo for most docker-related.

And the issue is with containerd.io (from docker repo) and not containerd (1.7.24~ds1-6+deb13u1 from debian repo). Adjusted the title and original post.

Sorry for the inaccuracy.