surajincloud
eks nodes fails with karpenter nodes
I haven't had time to troubleshoot this yet but I have a cluster with karpenter nodes
k get no
NAME STATUS ROLES AGE VERSION
i-001b401a8f4cb3b48.us-east-2.compute.internal Ready <none> 61m v1.27.3
i-002c826ec1bc8c7c9.us-east-2.compute.internal Ready <none> 53m v1.27.3
i-003ca3cd1e27c400c.us-east-2.compute.internal Ready <none> 63m v1.27.3
i-003f808cae84dd6cb.us-east-2.compute.internal Ready <none> 69m v1.27.3
i-004125fe7b0156226.us-east-2.compute.internal Ready <none> 45m v1.27.3
Trying to use the eks plugin gives me this error
k eks nodes
NAME INSTANCE-TYPE OS CAPACITY-TYPE ZONE AMI-ID AMI-NAME AGE
panic: runtime error: invalid memory address or nil pointer dereference
[signal SIGSEGV: segmentation violation code=0x1 addr=0x0 pc=0x3c739a6]
goroutine 1 [running]:
github.com/surajincloud/kubectl-eks/cmd.nodes(0x6193700?, {0x441a0f2?, 0x0?, 0x0?})
/home/runner/work/kubectl-eks/kubectl-eks/cmd/nodes.go:56 +0x886
github.com/spf13/cobra.(*Command).execute(0x6193700, {0x62d3ba8, 0x0, 0x0})
/home/runner/go/pkg/mod/github.com/spf13/[email protected]/command.go:916 +0x862
github.com/spf13/cobra.(*Command).ExecuteC(0x61939e0)
/home/runner/go/pkg/mod/github.com/spf13/[email protected]/command.go:1044 +0x3bd
github.com/spf13/cobra.(*Command).Execute(...)
/home/runner/go/pkg/mod/github.com/spf13/[email protected]/command.go:968
github.com/surajincloud/kubectl-eks/cmd.Execute()
/home/runner/work/kubectl-eks/kubectl-eks/cmd/root.go:25 +0x25
main.main()
/home/runner/work/kubectl-eks/kubectl-eks/main.go:6 +0x17
I have 2 nodes in a node group
k eks nodegroups
NAME RELEASE AMI_TYPE INSTANCE_TYPES STATUS
main 1.27.1-20230703 AL2_x86_64 t3.large ACTIVE
version
k eks version
v0.4.1
Hi @rothgar @sgreer81 I tried reproducing the error but I couldn't get the same error as you
kubectl eks version
v0.4.1
kubectl eks nodes
NAME INSTANCE-TYPE OS CAPACITY-TYPE ZONE AMI-ID AMI-NAME AGE
ip-192-168-152-204.eu-west-2.compute.internal t3.large linux ON_DEMAND eu-west-2c ami-09d73830fd8e6d486 amazon-eks-node-1.27-v20230825 16m
ip-192-168-168-19.eu-west-2.compute.internal t3.large linux ON_DEMAND eu-west-2b ami-09d73830fd8e6d486 amazon-eks-node-1.27-v20230825 16m
kubectl eks nodegroups
NAME RELEASE AMI_TYPE INSTANCE_TYPES STATUS
main 1.27.4-20230825 AL2_x86_64 t3.large ACTIVE
@rothgar @sgreer81 error is pointing to https://github.com/surajincloud/kubectl-eks/blob/main/cmd/nodes.go#L56. as per code, if there's no nodegroup image label it will look for karpenter image label on the node. I think your nodes don't have that label as well. and code is not really handling that error, i will get it fixed
fixed in this PR: https://github.com/surajincloud/kubectl-eks/pull/100
That doesn't crash for me although the AMI-NAME looks like it's using the ssm path
k eks nodes
NAME INSTANCE-TYPE OS CAPACITY-TYPE ZONE AMI-ID
AMI-NAME AGE
i-000a8e02bf3ff4081.us-east-2.compute.internal m5zn.6xlarge linux ON-DEMAND us-east-2c ami-000c38e2fb6e05120
ubuntu-eks/k8s_1.27/images/hvm-ssd/ubuntu-focal-20.04-amd64-server-20230816 6d21h
i-0052f4cf0b3d058c6.us-east-2.compute.internal c6a.12xlarge linux ON-DEMAND us-east-2a ami-000c38e2fb6e05120
ubuntu-eks/k8s_1.27/images/hvm-ssd/ubuntu-focal-20.04-amd64-server-20230816 6d21h
i-00ce125d01e31ffea.us-east-2.compute.internal c6a.4xlarge linux ON-DEMAND us-east-2c ami-000c38e2fb6e05120
ubuntu-eks/k8s_1.27/images/hvm-ssd/ubuntu-focal-20.04-amd64-server-20230816 6d21h
From your PR fix it looks like you could do another split on / and take [-1] to get the name. I have Amazon Linux nodes in the cluster too and they render properly in the output
@rothgar it's fixed now: https://github.com/surajincloud/kubectl-eks/pull/104