search

kubernetes-sigs / cluster-api-provider-nested

Cluster API Provider for Nested Clusters
Apache License 2.0
299 stars 65 forks source link

The pod is always pending in the nested cluster #156

Open jinsongo opened 3 years ago

jinsongo commented 3 years ago

What steps did you take and what happened: [A clear and concise description on how to REPRODUCE the bug.]

Follow the guide: https://github.com/kubernetes-sigs/cluster-api-provider-nested/blob/main/docs/README.md

  1. Deploy a nested cluster in Kind
  2. After nested cluster ready, deploy memcached service

The memcached pod is always pending, but the same deployment can work on real kubernetes environment

# kubectl --kubeconfig ./kubeconfig/kubeconfig.sample get pod
NAME          READY   STATUS    RESTARTS   AGE
memcached-0   0/1     Pending   0          6h43m
# kubectl --kubeconfig ./kubeconfig/kubeconfig.sample describe pod memcached-0
Name:           memcached-0
Namespace:      default
Priority:       0
Node:           <none>
Labels:         app=memcached
                controller-revision-hash=memcached-6b8cf9888
                statefulset.kubernetes.io/pod-name=memcached-0
Annotations:    <none>
Status:         Pending
IP:
IPs:            <none>
Controlled By:  StatefulSet/memcached
Containers:
  memcached-ct:
    Image:      memcached:1.5-alpine
    Port:       11211/TCP
    Host Port:  0/TCP
    Args:
      memcached
      -m
      256
    Environment:  <none>
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-zrvr9 (ro)
Volumes:
  default-token-zrvr9:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-zrvr9
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute op=Exists for 300s
                 node.kubernetes.io/unreachable:NoExecute op=Exists for 300s
Events:          <none>

What did you expect to happen: The memcatched pods are running, just like

# kubectl get pod
NAME                           READY   STATUS    RESTARTS   AGE
memcached-0                    1/1     Running   0          12m
memcached-1                    1/1     Running   0          12m

Anything else you would like to add: [Miscellaneous information that will assist in solving the issue.] The deploy yaml for memcached

apiVersion: v1
kind: Service
metadata:
  name: memcached
spec:
  ports:
  - port: 11211
  selector:
    app: memcached
  clusterIP: None
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
  name: memcached
spec:
  selector:
    matchLabels:
      app: memcached
  serviceName: "memcached"
  replicas: 2
  template:
    metadata:
      labels:
        app: memcached
    spec:
      restartPolicy: Always
      hostname: memcached
      containers:
      - name: memcached-ct
        image: memcached:1.5-alpine
        ports:
        - containerPort: 11211
        args: ["memcached", "-m", "256"]

Environment:

/kind bug [One or more /area label. See https://github.com/kubernetes-sigs/cluster-api-provider-nested/labels?q=area for the list of labels]

jinsongo commented 3 years ago

@Fei-Guo @christopherhein @charleszheng44 @gyliu513

christopherhein commented 3 years ago

Hey @wangjsty this is expected right now, we haven't finished the integration updates to immediately support CAPN + VC, the doc updates should be mostly called out in #141. Which is blocked on updating VC to support the way we're releasing images and manifests w/ CAPN/Prow.

Fei-Guo commented 3 years ago

Yes. At this moment, for VC demo, you can try the following demo which use the old clusterversion CR unless your purpose is to try CAPN specifically, https://github.com/kubernetes-sigs/cluster-api-provider-nested/blob/main/virtualcluster/doc/demo.md

jinsongo commented 3 years ago

@christopherhein @Fei-Guo Thank you, I will try the VC demo first, before the CAPN+VC integration done.

jinsongo commented 3 years ago

@christopherhein @Fei-Guo Could you please answer some questions for understanding. Thanks in advanced.

  1. Currently, the nested cluster that created by CAPN without VC can't be used to deploy workloads(deployment/statefulset) before CAPN+VC integration, is it right?
  2. Looks like it's okay to deploy workloads on VC without CAPN, what's the user scenario of CAPN? Or it could just be used as a common/unified cluster-API provider for VC? Thanks.
Fei-Guo commented 3 years ago

1) Yes 2) CAPN is the replacement for the original ClusterVersion CR/controller in VC. It follows CAPI standard and has much better manageability for tenant control plane Pods.

christopherhein commented 3 years ago

To add a little more color to #1 you could technically use this and run a set of data plane nodes w/ VMs or anything else and just use pod based control planes but you'd likely need to do some customizing as well, ie as of today that isn't supported.

jinsongo commented 3 years ago

@Fei-Guo @christopherhein Thank you for your explanation !

k8s-triage-robot commented 2 years ago

The Kubernetes project currently lacks enough contributors to adequately respond to all issues and PRs.

This bot triages issues and PRs according to the following rules:

You can:

Please send feedback to sig-contributor-experience at kubernetes/community.

/lifecycle stale

christopherhein commented 2 years ago

/remove-lifecycle stale /lifecycle frozen