Metrics server pods are not created on kOps

rodorgas

5/23/24, 7:49 PM

I'm using kOps 1.26.3 and Kubernetes 1.26.5, running on AWS. After upgrading with kops upgrade cluster, metrics server stopped working.

It is installed using this on cluster manifest:

  metricsServer:
    enabled: true

Some resources, like the service, are created. But there are no metric-server pods, nor deployments:

$ kubectl get service metrics-server -n kube-system
E0523 12:28:33.191542   35151 memcache.go:287] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:28:33.517819   35151 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:28:33.679998   35151 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:28:33.836750   35151 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
NAME             TYPE        CLUSTER-IP     EXTERNAL-IP   PORT(S)   AGE
metrics-server   ClusterIP   100.69.1.148   <none>        443/TCP   213d
$
$ kubectl get pods -A | grep metrics-server
E0523 12:44:00.464132   36275 memcache.go:287] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:44:00.780095   36275 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:44:00.942123   36275 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request
E0523 12:44:01.103146   36275 memcache.go:121] couldn't get resource list for metrics.k8s.io/v1beta1: the server is currently unable to handle the request

Since there are no pods, I don't know where to look for logs. I tried to terminate the instances and let kOps recreate them several times, with the same result.

244

0 + 2

kubernetes

kops

c4f4t0r

5/26/24, 12:54 PM

you have that error, because there is nod pod managing the request, you could check using "kubectl get apiservices | grep metrics" but I don't know why the pod metrics pod is crashing

rodorgas

5/31/24, 12:18 AM

@c4f4t0r: yeah it says the following in the column Available: "False (MissingEndpoints)". Its name is "v1beta1.metrics.k8s.io" and service is "kube-system/metrics-server".

Elon Musk

I sit in a Tesla and translated this thread with Ai:

EN: Metrics server pods are not created on kOps

Post an answer

Most people don’t grasp that asking a lot of questions unlocks learning and improves interpersonal bonding. In Alison’s studies, for example, though people could accurately recall how many questions had been asked in their conversations, they didn’t intuit the link between questions and liking. Across four studies, in which participants were engaged in conversations themselves or read transcripts of others’ conversations, people tended not to realize that question asking would influence—or had influenced—the level of amity between the conversationalists.