Failed scheduling

[cicdteam]

Observed this when vm failed to start as Pod was assigned to node but kubelet prevents it to start

Events:
  Type     Reason            Age   From                 Message
  ----     ------            ----  ----                 -------
  Warning  FailedScheduling  43m   autoscale-scheduler  0/3 nodes are available: 1 Insufficient neonvm/kvm, 3 Insufficient cpu.
  Warning  FailedScheduling  43m   autoscale-scheduler  0/3 nodes are available: 1 Insufficient neonvm/kvm, 3 Insufficient cpu.
  Normal   Scheduled         43m   autoscale-scheduler  Successfully assigned default/vm-stress-0050-dz557 to i-08d5967318f466dd4.eu-west-1.compute.internal
  Warning  OutOfcpu          43m   kubelet              Node didn't have enough resource: cpu, requested: 1000, used: 94939, capacity: 95690

As result VM stuck in failed state.

vm

% k get neonvm vm-stress-0050      
NAME             CPUS   MEMORY   POD                    EXTRAIP         STATUS   AGE
vm-stress-0050                   vm-stress-0050-dz557   10.100.128.49   Failed   48m

pod

% k get po vm-stress-0050-dz557
NAME                   READY   STATUS     RESTARTS   AGE
vm-stress-0050-dz557   0/1     OutOfcpu   0          48m

pod status

% k get po vm-stress-0050-dz557 -ojson| jq '.status'  
{
  "message": "Pod Node didn't have enough resource: cpu, requested: 1000, used: 94939, capacity: 95690",
  "phase": "Failed",
  "reason": "OutOfcpu",
  "startTime": "2023-07-04T11:19:28Z"
}