我一直试图在Kubernetes上运行kafka / zookeeper。使用舵图，我可以在集群上安装zookeeper。但是，ZK吊舱卡在挂起状态。当我在其中一个Pod“ didn't find available persistent volumes to bind, 1 node(s) had taints that the pod didn't tolerate.”上发布describe时，是调度失败的原因。但是，当我在PVC上发布describe时，得到的是“ waiting for first consumer to be created before binding”。我试图重新生成整个集群，但是结果是一样的。尝试使用https://kubernetes.io/blog/2018/04/13/local-persistent-volumes-beta/作为指导。
有人可以在这里引导我吗？
kubectl获得豆荚-n动物园管理员
kubectl get pods -n zoo-keeper
NAME READY STATUS RESTARTS AGE
zoo-keeper-zk-0 0/1 Pending 0 20m
zoo-keeper-zk-1 0/1 Pending 0 20m
zoo-keeper-zk-2 0/1 Pending 0 20m
kubectl获取sc
kubectl get sc
NAME PROVISIONER AGE
local-storage kubernetes.io/no-provisioner 25m
kubectl描述sc
kubectl describe sc
Name: local-storage
IsDefaultClass: No
Annotations: kubectl.kubernetes.io/last-applied-configuration={"apiVersion":"storage.k8s.io/v1","kind":"StorageClass","metadata":{"annotations":{},"name":"local-storage"},"provisioner":"kubernetes.io/no-provisioner","volumeBindingMode":"WaitForFirstConsumer"}
Provisioner: kubernetes.io/no-provisioner
Parameters: <none>
AllowVolumeExpansion: <unset>
MountOptions: <none>
ReclaimPolicy: Delete
VolumeBindingMode: WaitForFirstConsumer
Events: <none>
kubectl描述pod foob-zookeeper-0 -n zookeeper
ubuntu@kmaster:~$ kubectl describe pod foob-zookeeper-0 -n zoo-keeper
Name: foob-zookeeper-0
Namespace: zoo-keeper
Priority: 0
PriorityClassName: <none>
Node: <none>
Labels: app=foob-zookeeper
app.kubernetes.io/instance=data-coord
app.kubernetes.io/managed-by=Tiller
app.kubernetes.io/name=foob-zookeeper
app.kubernetes.io/version=foob-zookeeper-9.1.0-15
controller-revision-hash=foob-zookeeper-5321f8ff5
release=data-coord
statefulset.kubernetes.io/pod-name=foob-zookeeper-0
Annotations: foobar.com/product-name: zoo-keeper ZK
foobar.com/product-revision: ABC
Status: Pending
IP:
Controlled By: StatefulSet/foob-zookeeper
Containers:
foob-zookeeper:
Image: repo.data.foobar.se/latest/zookeeper-3.4.10:1.6.0-15
Ports: 2181/TCP, 2888/TCP, 3888/TCP, 10007/TCP
Host Ports: 0/TCP, 0/TCP, 0/TCP, 0/TCP
Limits:
cpu: 2
memory: 4Gi
Requests:
cpu: 1
memory: 2Gi
Liveness: exec [zkOk.sh] delay=15s timeout=5s period=10s #success=1 #failure=3
Readiness: tcp-socket :2181 delay=15s timeout=5s period=10s #success=1 #failure=3
Environment:
ZK_REPLICAS: 3
ZK_HEAP_SIZE: 1G
ZK_TICK_TIME: 2000
ZK_INIT_LIMIT: 10
ZK_SYNC_LIMIT: 5
ZK_MAX_CLIENT_CNXNS: 60
ZK_SNAP_RETAIN_COUNT: 3
ZK_PURGE_INTERVAL: 1
ZK_LOG_LEVEL: INFO
ZK_CLIENT_PORT: 2181
ZK_SERVER_PORT: 2888
ZK_ELECTION_PORT: 3888
JMXPORT: 10007
Mounts:
/var/lib/zookeeper from datadir (rw)
/var/run/secrets/kubernetes.io/serviceaccount from default-token-nfcfx (ro)
Conditions:
Type Status
PodScheduled False
Volumes:
datadir:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: datadir-foob-zookeeper-0
ReadOnly: false
default-token-nfcfx:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-nfcfx
Optional: false
QoS Class: Burstable
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 69s (x4 over 3m50s) default-scheduler 0/2 nodes are available: 1 node(s) didn't find available persistent volumes to bind, 1 node(s) had taints that the pod didn't tolerate.
kubectl获得pv
ubuntu@kmaster:~$ kubectl get pv
NAME CAPACITY ACCESS MODES RECLAIM POLICY STATUS CLAIM STORAGECLASS REASON AGE
local-pv 50Gi RWO Retain Available local-storage 10m
ubuntu@kmaster:~$
kubectl获得pvc本地声明
ubuntu@kmaster:~$ kubectl get pvc local-claim
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
local-claim Pending local-storage 8m9s
ubuntu@kmaster:~$
kubectl描述pvc本地声明
ubuntu@kmaster:~$ kubectl describe pvc local-claim
Name: local-claim
Namespace: default
StorageClass: local-storage
Status: Pending
Volume:
Labels: <none>
Annotations: <none>
Finalizers: [kubernetes.io/pvc-protection]
Capacity:
Access Modes:
VolumeMode: Filesystem
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Normal WaitForFirstConsumer 2m3s (x26 over 7m51s) persistentvolume-controller waiting for first consumer to be created before binding
Mounted By: <none>
我的PV文件：
cat create-pv.yml
apiVersion: v1
kind: PersistentVolume
metadata:
name: local-pv
spec:
capacity:
storage: 50Gi
accessModes:
- ReadWriteOnce
persistentVolumeReclaimPolicy: Retain
storageClassName: local-storage
local:
path: /mnt/kafka-mount
nodeAffinity:
required:
nodeSelectorTerms:
- matchExpressions:
- key: kubernetes.io/hostname
operator: In
values:
- kmaster
猫pvc.yml
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
name: local-claim
spec:
accessModes:
- ReadWriteOnce
storageClassName: local-storage
resources:
requests:
storage: 50Gi
答案 0 :(得分：3)
好像您在主节点上创建了PV。默认情况下，普通节点使用所谓的taint将主节点标记为不可调度。为了能够在主节点上运行某些服务，您有两个选择：
1）为某些服务添加容忍度，以使其可以在主节点上运行：
tolerations:
- effect: NoSchedule
key: node-role.kubernetes.io/master
您甚至可以指定某些服务仅在主节点上运行：
nodeSelector:
node-role.kubernetes.io/master: ""
2）您可以从主节点上删除污点，以便任何pod都可以在其上运行。您应该知道这很危险，因为它会使群集非常不稳定。
kubectl taint nodes --all node-role.kubernetes.io/master-
在此处了解更多信息以及污点和公差：https://kubernetes.io/docs/concepts/configuration/taint-and-toleration/