Skip to content

aws-cni Failed to create pod sandbox error #3410

@nikhlagrwl

Description

@nikhlagrwl

What happened:
We have an EKS cluster in AWS with self managed Linux node groups, instance type - c5a.2xlarge. We have configured warm pool (stopped state) in the Auto Scaling Group and whenever any pods are coming up which is scaling up instances from warm pool to normal pool, we are getting the below logs in the cluster events and the overall pod start time is increasing.

Warning FailedCreatePodSandBox 118s kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to setup network for sandbox "719d35bdbfe36a98c614414306a9596bd9fa1c8a1c5db554b2ed19a44cb54b04": plugin type="aws-cni" name="aws-cni" failed (add): add cmd: Error received from AddNetwork gRPC call: rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 127.0.0.1:50051: connect: connection refused"

Attach logs
Warning FailedCreatePodSandBox 118s kubelet Failed to create pod sandbox: rpc error: code = Unknown desc = failed to setup network for sandbox "719d35bdbfe36a98c614414306a9596bd9fa1c8a1c5db554b2ed19a44cb54b04": plugin type="aws-cni" name="aws-cni" failed (add): add cmd: Error received from AddNetwork gRPC call: rpc error: code = Unavailable desc = connection error: desc = "transport: Error while dialing: dial tcp 127.0.0.1:50051: connect: connection refused"

What you expected to happen:

How to reproduce it (as minimally and precisely as possible):

Anything else we need to know?:

Environment:

  • Kubernetes version (use kubectl version):
  • CNI Version: v1.19.2-eksbuild.5
  • OS (e.g: cat /etc/os-release):
  • Kernel (e.g. uname -a):

Metadata

Metadata

Assignees

No one assigned

    Labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions