Unable to set limit in K8 Multi node playground?

Kubernetes
, , , ,
#1

I am using following config file to set resource Quotas:

apiVersion: v1
kind: Pod
metadata:
  labels:
    run: kplabs-quota
  name: kplabs-quota
spec:
  containers:
  - image: nginx
    name: kplabs-quota
    resources:
      requests:
        memory: "128Mi"
      limits:
        memory: "512Mi"
        cpu: "2"

But it is giving following error:
controlplane ~ ➜ k get pods
NAME READY STATUS RESTARTS AGE
kplabs-quota 0/1 ContainerCreating 0 3s

controlplane ~ ➜ k get pods
NAME READY STATUS RESTARTS AGE
kplabs-quota 0/1 CrashLoopBackOff 8 (3m37s ago) 19m

k describe kplabs-quota is giving me following error:
Message:      failed to create containerd task: failed to create shim task: OCI runtime create failed: runc create failed: unable to start container process: error during container init: error setting cgroup config for procHooks process: failed to write "200000": write /sys/fs/cgroup/cpu,cpuacct/system.slice/containerd.service/kubepods-burstable-podd6e08c5a_0269_4b6e_ba59_e25d23ef5886.slice:cri-containerd:3231c97a0ca7925b34c5ec73ce04cf9e3d4ac0301d219283da7b4e65fc259445/cpu.cfs_quota_us: invalid argument: unknown

I am running this in Kodekloud playground latest multi node

controlplane ~ ✖ kubectl describe node node01 | grep Allocatable -A 5
Allocatable:
  cpu:                18
  ephemeral-storage:  936398358207
  hugepages-1Gi:      0
  hugepages-2Mi:      0
  memory:             98785908Ki
--
  Normal   NodeAllocatableEnforced  33m                kubelet          Updated Node Allocatable limit across pods
  Normal   RegisteredNode           32m                node-controller  Node node01 event: Registered Node node01 in Controller
  Normal   NodeReady                32m                kubelet          Node node01 status is now: NodeReady

Same for node02 and controlplane

controlplane ~ ➜  uname -r
5.4.0-1106-gcp
#2

Not sure why you’re getting the error, but it may be that you are asking for more cpu than the PG will grant you. I commented out the resources.limits.cpu line, and the YAML then worked.