附加新卷（EKS）时Kubernetes Pod处于挂起状态

Question

让我描述一下我的情况：

TL; DR

当我在带有1个附加卷的Kubernetes上创建部署时，一切工作正常。当我创建相同的部署，但附加了第二个卷（总计：2个卷）时，吊舱因错误而卡在“待定”上：

pod has unbound PersistentVolumeClaims (repeated 2 times)
0/2 nodes are available: 2 node(s) had no available volume zone.

已检查是否在正确的可用区中创建了卷。

详细说明

我有一个使用2个节点的Amazon EKS设置的集群。我有以下默认存储类：

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: gp2
  annotations:
    storageclass.kubernetes.io/is-default-class: "true"
provisioner: kubernetes.io/aws-ebs
parameters:
  type: gp2
reclaimPolicy: Retain
mountOptions:
  - debug

我有一个mongodb部署，它需要两个卷，一个卷安装在/data/db文件夹上，另一个卷安装在我需要的某个随机目录中。这是用于创建三个组件的最小Yaml（我故意评论了几行）：

apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  namespace: my-project
  creationTimestamp: null
  labels:
    io.kompose.service: my-project-db-claim0
  name: my-project-db-claim0
spec:
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 5Gi
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
  namespace: my-project
  creationTimestamp: null
  labels:
    io.kompose.service: my-project-db-claim1
  name: my-project-db-claim1
spec:
  accessModes:
  - ReadWriteOnce
  resources:
    requests:
      storage: 10Gi
---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  namespace: my-project
  name: my-project-db
spec:
  replicas: 1
  strategy:
    type: Recreate
  template:
    metadata:
      labels:
        name: my-db
    spec:
      containers:
        - name: my-project-db-container
          image: mongo
          imagePullPolicy: Always
          resources: {}
          volumeMounts:
          - mountPath: /my_dir
            name: my-project-db-claim0
          # - mountPath: /data/db
          #   name: my-project-db-claim1
          ports:
            - containerPort: 27017
      restartPolicy: Always
      volumes:
      - name: my-project-db-claim0
        persistentVolumeClaim:
          claimName: my-project-db-claim0
      # - name: my-project-db-claim1
      #   persistentVolumeClaim:
      #     claimName: my-project-db-claim1

该yaml完美地工作。卷的输出为：

$ kubectl describe pv

Name:            pvc-307b755a-039e-11e9-b78d-0a68bcb24bc6
Labels:          failure-domain.beta.kubernetes.io/region=us-east-1
                failure-domain.beta.kubernetes.io/zone=us-east-1c
Annotations:     kubernetes.io/createdby: aws-ebs-dynamic-provisioner
                pv.kubernetes.io/bound-by-controller: yes
                pv.kubernetes.io/provisioned-by: kubernetes.io/aws-ebs
Finalizers:      [kubernetes.io/pv-protection]
StorageClass:    gp2
Status:          Bound
Claim:           my-project/my-project-db-claim0
Reclaim Policy:  Delete
Access Modes:    RWO
Capacity:        5Gi
Node Affinity:   <none>
Message:        
Source:
    Type:       AWSElasticBlockStore (a Persistent Disk resource in AWS)
    VolumeID:   aws://us-east-1c/vol-xxxxx
    FSType:     ext4
    Partition:  0
    ReadOnly:   false
Events:         <none>


Name:            pvc-308d8979-039e-11e9-b78d-0a68bcb24bc6
Labels:          failure-domain.beta.kubernetes.io/region=us-east-1
                failure-domain.beta.kubernetes.io/zone=us-east-1b
Annotations:     kubernetes.io/createdby: aws-ebs-dynamic-provisioner
                pv.kubernetes.io/bound-by-controller: yes
                pv.kubernetes.io/provisioned-by: kubernetes.io/aws-ebs
Finalizers:      [kubernetes.io/pv-protection]
StorageClass:    gp2
Status:          Bound
Claim:           my-project/my-project-db-claim1
Reclaim Policy:  Delete
Access Modes:    RWO
Capacity:        10Gi
Node Affinity:   <none>
Message:        
Source:
    Type:       AWSElasticBlockStore (a Persistent Disk resource in AWS)
    VolumeID:   aws://us-east-1b/vol-xxxxx
    FSType:     ext4
    Partition:  0
    ReadOnly:   false
Events:         <none>

和pod输出：

$ kubectl describe pods

Name:               my-project-db-7d48567b48-slncd
Namespace:          my-project
Priority:           0
PriorityClassName:  <none>
Node:               ip-192-168-212-194.ec2.internal/192.168.212.194
Start Time:         Wed, 19 Dec 2018 15:55:58 +0100
Labels:             name=my-db
                    pod-template-hash=3804123604
Annotations:        <none>
Status:             Running
IP:                 192.168.216.33
Controlled By:      ReplicaSet/my-project-db-7d48567b48
Containers:
  my-project-db-container:
    Container ID:   docker://cf8222f15e395b02805c628b6addde2d77de2245aed9406a48c7c6f4dccefd4e
    Image:          mongo
    Image ID:       docker-pullable://mongo@sha256:0823cc2000223420f88b20d5e19e6bc252fa328c30d8261070e4645b02183c6a
    Port:           27017/TCP
    Host Port:      0/TCP
    State:          Running
      Started:      Wed, 19 Dec 2018 15:56:42 +0100
    Ready:          True
    Restart Count:  0
    Environment:    <none>
    Mounts:
      /my_dir from my-project-db-claim0 (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-pf9ks (ro)
Conditions:
  Type           Status
  Initialized    True
  Ready          True
  PodScheduled   True
Volumes:
  my-project-db-claim0:
    Type:       PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
    ClaimName:  my-project-db-claim0
    ReadOnly:   false
  default-token-pf9ks:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-pf9ks
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason                  Age                    From                                      Message
  ----     ------                  ----                   ----                                      -------
  Warning  FailedScheduling        7m22s (x5 over 7m23s)  default-scheduler                         pod has unbound PersistentVolumeClaims (repeated 2 times)
  Normal   Scheduled               7m21s                  default-scheduler                         Successfully assigned my-project/my-project-db-7d48567b48-slncd to ip-192-168-212-194.ec2.internal
  Normal   SuccessfulMountVolume   7m21s                  kubelet, ip-192-168-212-194.ec2.internal  MountVolume.SetUp succeeded for volume "default-token-pf9ks"
  Warning  FailedAttachVolume      7m13s (x5 over 7m21s)  attachdetach-controller                   AttachVolume.Attach failed for volume "pvc-307b755a-039e-11e9-b78d-0a68bcb24bc6" : "Error attaching EBS volume \"vol-01a863d0aa7c7e342\"" to instance "i-0a7dafbbdfeabc50b" since volume is in "creating" state
  Normal   SuccessfulAttachVolume  7m1s                   attachdetach-controller                   AttachVolume.Attach succeeded for volume "pvc-307b755a-039e-11e9-b78d-0a68bcb24bc6"
  Normal   SuccessfulMountVolume   6m48s                  kubelet, ip-192-168-212-194.ec2.internal  MountVolume.SetUp succeeded for volume "pvc-307b755a-039e-11e9-b78d-0a68bcb24bc6"
  Normal   Pulling                 6m48s                  kubelet, ip-192-168-212-194.ec2.internal  pulling image "mongo"
  Normal   Pulled                  6m39s                  kubelet, ip-192-168-212-194.ec2.internal  Successfully pulled image "mongo"
  Normal   Created                 6m38s                  kubelet, ip-192-168-212-194.ec2.internal  Created container
  Normal   Started                 6m37s                  kubelet, ip-192-168-212-194.ec2.internal  Started container

一切都已创建，没有任何问题。但是，如果我取消注释yaml中的行，因此将两个卷连接到db部署，则pv输出与之前的输出相同，但是pod停留在挂起状态，并显示以下输出：

$ kubectl describe pods

Name:               my-project-db-b8b8d8bcb-l64d7
Namespace:          my-project
Priority:           0
PriorityClassName:  <none>
Node:               <none>
Labels:             name=my-db
                    pod-template-hash=646484676
Annotations:        <none>
Status:             Pending
IP:                 
Controlled By:      ReplicaSet/my-project-db-b8b8d8bcb
Containers:
  my-project-db-container:
    Image:        mongo
    Port:         27017/TCP
    Host Port:    0/TCP
    Environment:  <none>
    Mounts:
      /data/db from my-project-db-claim1 (rw)
      /my_dir from my-project-db-claim0 (rw)
      /var/run/secrets/kubernetes.io/serviceaccount from default-token-pf9ks (ro)
Conditions:
  Type           Status
  PodScheduled   False 
Volumes:
  my-project-db-claim0:
    Type:       PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
    ClaimName:  my-project-db-claim0
    ReadOnly:   false
  my-project-db-claim1:
    Type:       PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
    ClaimName:  my-project-db-claim1
    ReadOnly:   false
  default-token-pf9ks:
    Type:        Secret (a volume populated by a Secret)
    SecretName:  default-token-pf9ks
    Optional:    false
QoS Class:       BestEffort
Node-Selectors:  <none>
Tolerations:     node.kubernetes.io/not-ready:NoExecute for 300s
                node.kubernetes.io/unreachable:NoExecute for 300s
Events:
  Type     Reason            Age                From               Message
  ----     ------            ----               ----               -------
  Warning  FailedScheduling  60s (x5 over 60s)  default-scheduler  pod has unbound PersistentVolumeClaims (repeated 2 times)
  Warning  FailedScheduling  2s (x16 over 59s)  default-scheduler  0/2 nodes are available: 2 node(s) had no available volume zone.

我已经阅读了以下两个问题：

动态卷配置在错误的可用区域中创建EBS卷

可以在没有节点的可用性区域中创建EBS上的PersistentVolume（已关闭）

但是我已经检查了卷是否在与群集节点实例相同的区域中创建。实际上，EKS默认在us-east-1b和us-east-1c区域中创建两个EBS，并且这些卷有效。由发布的Yaml创建的卷也在这些区域上。

Answer 1

听起来它正在尝试在没有任何卷的可用区域中创建卷。您可以尝试将您的节点限制StorageClass在您拥有节点的可用区域。

kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: gp2
  annotations:
    storageclass.kubernetes.io/is-default-class: "true"
provisioner: kubernetes.io/aws-ebs
parameters:
  type: gp2
reclaimPolicy: Retain
mountOptions:
  - debug
allowedTopologies:
- matchLabelExpressions:
  - key: failure-domain.beta.kubernetes.io/zone
    values:
    - us-east-1b
    - us-east-1c

这与这个问题和这个答案非常相似，只是描述的问题是在 GCP 上，在本例中是 AWS。