[ws-daemon] cannot set CFS quota

[sagor999]

Bug description

seeing this in log spamming non stop:

"cannot set CFS quota:
    github.com/gitpod-io/gitpod/ws-daemon/pkg/resources.cgroupCFSController.SetQuota
        github.com/gitpod-io/gitpod/ws-daemon/pkg/resources/controller.go:420
  - write /mnt/node-cgroups/cpu/kubepods/burstable/pod4c349442-a7f3-401b-a667-4bc68ee5ba45/f87cae5fd4e0fb23571a777086c3be55f564cabdf3fe256e650aafdf9a2b589e/cpu.cfs_quota_us: invalid argument"

might be related to this: #5326

customer reported following:

One of my workspaces times out on startup after 3-4 minutes with the following error message:

Oh, no! Something went wrong!
Error: no available workspace cluster to choose from!

(This is harlequin-limpet-1t9g3dau.). The error has been going for the last 45 mins or so, and I've noticed increasing intermittent errors over the last few days.

Not sure if timeout related to error that I am seeing in logs.

Steps to reproduce

search for affected workspace in logs.

Workspace affected

harlequin-limpet-1t9g3dau

Expected behavior

No response

Example repository

No response

Anything else?

No response

[sagor999]

@csweichel do you know what might be going on here?

[aledbf]

@sagor999 please continue working on #7797

[sagor999]

This should help to get to the bottom of this once it is rolled out to production:
#8390

[Furisto]

I do not see this in the logs for eu34 and us34.

[sagor999]

Yeap, seems like this was fixed now.