MLRun deploy, 0/3 nodes are available: 3 Insufficient cpu

I got this issue during deploy of python code in MLRun, see build log:

Failed to wait for function resources to be available
   .../platform/kube/controller/nucliofunction.go:230
   
   ...//nuclio/pkg/platform/kube/client/deployer.go:310
Failed to wait for function readiness.

Pod logs:

* nuclio-risk-behcol-behcol-scoring-57890098-kln2h

* Warning events: 
0/3 nodes are available: 3 Insufficient cpu.

    ...//nuclio/pkg/platform/kube/client/deployer.go:140

Do you know, why I got this issues and how can I avoid it?

>Solution :

It seems, that your total CPU requests/limits on your K8s are out of available sources.

You have two options:

  • You have to decrease CPU requests/limits for existing pods or
  • You have to increase amount of CPUs in your K8s

Leave a Reply