network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address

5/26/2019

What I try to do: EKS with both Linux and Windows (2019) nodes, nginx pod on Linux should access IIS pod on Windows.

The issue: The Windows pods don't start.

Log:

E0526 10:59:31.963644    4392 pod_workers.go:186] Error syncing pod b35e92cc-7fa2-11e9-b07b-0ac0c740dc70 ("phoenix-57b76c578c-cczs2_kaltura(b35e92cc-7fa2-11e9-b07b-0ac0c740dc70)"), skipping: failed to "KillPodSandbox" for "b35e92cc-7fa2-11e9-b07b-0ac0c740dc70" with KillPodSandboxError: "rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod \"phoenix-57b76c578c-cczs2_kaltura\" network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address"
I0526 10:59:37.049583    5020 proxier.go:117] Hns Endpoint resource, {"ID":"9638A3AE-DCB9-4F85-B682-9D2879E09D98","Name":"Ethernet","VirtualNetwork":"82363D68-76A8-4225-8EFC-76F179330CC1","VirtualNetworkName":"vpcbr0a05d9b85b68","Policies":[{"Type":"L2Driver"}],"MacAddress":"00:11:22:33:44:55","IPAddress":"172.31.32.190","PrefixLength":20,"IsRemoteEndpoint":true}
I0526 10:59:37.051589    5020 proxier.go:117] Hns Endpoint resource, {"ID":"8A4C02B1-537B-4650-ADC5-BA24598E3ABA","Name":"Ethernet","VirtualNetwork":"82363D68-76A8-4225-8EFC-76F179330CC1","VirtualNetworkName":"vpcbr0a05d9b85b68","Policies":[{"Type":"L2Driver"}],"MacAddress":"00:11:22:33:44:55","IPAddress":"172.31.36.90","PrefixLength":20,"IsRemoteEndpoint":true}
E0526 10:59:37.064582    5020 proxier.go:1034] Policy creation failed: hnsCall failed in Win32: The provided policy configuration is invalid or missing parameters. (0x803b000d)
E0526 10:59:37.064582    5020 proxier.go:1018] Endpoint information not available for service kaltura/phoenix:https. Not applying any policy
E0526 10:59:38.433836    4392 kubelet_network.go:102] Failed to ensure that nat chain KUBE-MARK-DROP exists: error creating chain "KUBE-MARK-DROP": executable file not found in %PATH%: 
E0526 10:59:39.362013    4392 helpers.go:735] eviction manager: failed to construct signal: "allocatableMemory.available" error: system container "pods" not found in metrics
W0526 10:59:39.362013    4392 helpers.go:808] eviction manager: no observation found for eviction signal nodefs.inodesFree
E0526 10:59:48.965710    4392 cni.go:280] Error deleting network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address
E0526 10:59:48.965710    4392 remote_runtime.go:115] StopPodSandbox "04961285217a628c589467359f6ff6335355c73fdd61f3c975215105a6c307f6" from runtime service failed: rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod "phoenix-57b76c578c-cczs2_kaltura" network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address
E0526 10:59:48.965710    4392 kuberuntime_manager.go:799] Failed to stop sandbox {"docker" "04961285217a628c589467359f6ff6335355c73fdd61f3c975215105a6c307f6"}
E0526 10:59:48.965710    4392 kuberuntime_manager.go:594] killPodWithSyncResult failed: failed to "KillPodSandbox" for "b35e92cc-7fa2-11e9-b07b-0ac0c740dc70" with KillPodSandboxError: "rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod \"phoenix-57b76c578c-cczs2_kaltura\" network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address"
E0526 10:59:48.965710    4392 pod_workers.go:186] Error syncing pod b35e92cc-7fa2-11e9-b07b-0ac0c740dc70 ("phoenix-57b76c578c-cczs2_kaltura(b35e92cc-7fa2-11e9-b07b-0ac0c740dc70)"), skipping: failed to "KillPodSandbox" for "b35e92cc-7fa2-11e9-b07b-0ac0c740dc70" with KillPodSandboxError: "rpc error: code = Unknown desc = NetworkPlugin cni failed to teardown pod \"phoenix-57b76c578c-cczs2_kaltura\" network: failed to parse Kubernetes args: pod does not have label vpc.amazonaws.com/PrivateIPv4Address"
E0526 10:59:49.368785    4392 helpers.go:735] eviction manager: failed to construct signal: "allocatableMemory.available" error: system container "pods" not found in metrics
W0526 10:59:49.368785    4392 helpers.go:808] eviction manager: no observation found for eviction signal nodefs.inodesFree

kubectl -n kaltura describe pods phoenix-695b5bdff8-zzbq6

Name:               phoenix-695b5bdff8-zzbq6
Namespace:          kaltura
Priority:           0
PriorityClassName:  <none>
Node:               ip-10-10-12-97.us-east-2.compute.internal/10.10.12.97
Start Time:         Tue, 28 May 2019 12:30:48 +0300
Labels:             app.kubernetes.io/instance=kaltura-core
                    app.kubernetes.io/name=phoenix
                    pod-template-hash=2516168994
Annotations:        <none>
Status:             Pending
IP:
Controlled By:      ReplicaSet/phoenix-695b5bdff8
Containers:
  kaltura:
    Container ID:
    Image:          <my-account-id>.dkr.ecr.us-east-2.amazonaws.com/vfd1-phoenix:latest
    Image ID:
    Port:           8040/TCP
    Host Port:      0/TCP
    State:          Waiting
      Reason:       ContainerCreating
    Ready:          False
    Restart Count:  0
    Liveness:       http-get http://:80/tvp_api delay=0s timeout=1s period=10s #success=1 #failure=3
    Readiness:      http-get http://:80/tvp_api delay=0s timeout=1s period=10s #success=1 #failure=3
    Environment:
      TCM_SECTION:  kaltura-core
      TCM_URL:      https://10.10.12.99
    Mounts:
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-jdd98 (ro)
Conditions:
  Type              Status
  Initialized       True
  Ready             False
  ContainersReady   False
  PodScheduled      True
Volumes:
  default-token-jdd98:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-jdd98
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  beta.kubernetes.io/arch=amd64
                 beta.kubernetes.io/os=windows
                 kaltura.role=api
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                 node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type    Reason          Age                      From                                                Message
  ----    ------          ----                     ----                                                -------
  Normal  SandboxChanged  113s (x1707 over 7h27m)  kubelet, ip-10-10-12-97.us-east-2.compute.internal  Pod sandbox changed, it will be killed and re-created.

Deployment yaml (from helm):

apiVersion: v1
kind: Service
metadata:
  name: phoenix
  labels:
    app.kubernetes.io/name: phoenix
    app.kubernetes.io/instance: {{ .Release.Name }}
    app.kubernetes.io/managed-by: {{ .Release.Service }}
spec:
  type: ClusterIP
  ports:
    - port: 8080
      targetPort: 443
      protocol: TCP
      name: https
  selector:
    app.kubernetes.io/name: phoenix
    app.kubernetes.io/instance: {{ .Release.Name }}


---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: phoenix
  labels:
    app.kubernetes.io/name: phoenix
    app.kubernetes.io/instance: {{ .Release.Name }}
    app.kubernetes.io/managed-by: {{ .Release.Service }}
spec:
  replicas: 2
  selector:
    matchLabels:
      app.kubernetes.io/name: phoenix
      app.kubernetes.io/instance: {{ .Release.Name }}
  template:
    metadata:
      labels:
        app.kubernetes.io/name: phoenix
        app.kubernetes.io/instance: {{ .Release.Name }}
    spec:
      containers:
        - name: {{ .Chart.Name }}
          image: "{{ .Values.env.repository }}/{{ .Values.env.tag }}-phoenix:latest"
          imagePullPolicy: Always
          env:
            - name: TCM_SECTION
              value: {{ .Values.env.tag }}
          ports:
            - name: http
              containerPort: 8040
              protocol: TCP
          livenessProbe:
            httpGet:
              path: /tvp_api
              port: 80
          readinessProbe:
            httpGet:
              path: /tvp_api
              port: 80
      strategy:
        type: RollingUpdate
        maxUnavailable: 1
      nodeSelector:
        kaltura.role: api
        beta.kubernetes.io/os: windows
        beta.kubernetes.io/arch: amd64

Additionally to this pod I have an nginx pod running on linux nodes, that pod is load-balanced using aws-alb-ingress-controller.

-- Johnathan Kanarek
eks
kubernetes
windows

1 Answer

6/30/2019

Solved. Apparently the vpc admission webhook was defined on the default namespace while my deployment of windows pods was on a different namespace.

-- Johnathan Kanarek
Source: StackOverflow