"rep" fails on Ubuntu Noble stemcell

[jochenehret]

Current behavior

We have started to evaluate the Ubuntu "noble" stemcell with this validation pipeline:
https://concourse.wg-ard.ci.cloudfoundry.org/teams/main/pipelines/noble-stemcell-validation
Currently the "rep" job fails to start. In the rep.stdout.log file we see this error message:

{"timestamp":"2024-12-09T11:39:22.423211303Z","level":"error","source":"rep","message":"rep.failed-to-initialize-executor","data":{"error":"failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error mounting \"cgroup\" to rootfs at \"/sys/fs/cgroup\": mount /sys/fs/cgroup:/sys/fs/cgroup (via /proc/self/fd/9), flags: 0x100f: invalid argument: unknown"}}

Stemcell version is bosh-google-kvm-ubuntu-noble / 0.26. "diego-release" version is 2.108.0.

Desired behavior

The "rep" job should start successfully.

Affected Version

2.108.0

[winkingturtle-vmw]

Yeah, I think that would be expected since guardian doesn't have support for cgroups v2. Assigning to @mariash since she has been working on this.

[oliver-heinrich]

Hi @mariash, I'm from the CF-deployment team. First of all happy new year :)
Is there meanwhile any update from your side?

[mariash]

Hi @oliver-heinrich,

garden-runc was just release v1.67.0 with cgroups v2 support. It should work with Noble stemcell. Let us know how it works for you.

[oliver-heinrich]

Hi @mariash , awesome thank you for the info. @dimivel can you keep Maria posted about the results when we rerun the Noble validation pipeline?

[jochenehret]

The deployments now succeeds with garden-runc/1.67.0:
https://concourse.wg-ard.ci.cloudfoundry.org/teams/main/pipelines/noble-stemcell-validation/jobs/deploy-cf/builds/18

Closing this issue, thanks for your support!