k8s使用ceph實現動態持久化存儲

簡介

本文章介紹如何使用ceph爲k8s提供動態申請pv的功能。ceph提供底層存儲功能，cephfs方式支持k8s的pv的3種訪問模式ReadWriteOnce，ReadOnlyMany ，ReadWriteMany ，RBD支持ReadWriteOnce，ReadOnlyMany兩種模式

訪問模式只是能力描述，並非強制執行的，對於沒有按pvc聲明的方式使用pv，存儲提供者應該負責訪問時的運行錯誤。例如若是設置pvc的訪問模式爲ReadOnlyMany ，pod掛載後依然可寫，若是須要真正的不可寫，申請pvc是須要指定 readOnly: true 參數

部署

部署k8s

centos7使用kubeadm安裝k8s-1.11版本

部署ceph

centos7安裝ceph分佈式存儲集羣

在k8s集羣中配置使用ceph

使用Ceph RBD

使用kubeadm安裝集羣的額外配置

# 若是使用kubeadm部署的集羣須要這些額外的步驟
# 因爲使用動態存儲時 controller-manager 須要使用 rbd 命令建立 image
# 因此 controller-manager 須要使用 rbd 命令
# 因爲官方controller-manager鏡像裏沒有rbd命令
# 若是沒使用以下方式會報錯沒法成功建立pvc
# 相關 issue https://github.com/kubernetes/kubernetes/issues/38923
cat >external-storage-rbd-provisioner.yaml<<EOF
apiVersion: v1
kind: ServiceAccount
metadata:
  name: rbd-provisioner
  namespace: kube-system
---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: rbd-provisioner
rules:
  - apiGroups: [""]
    resources: ["persistentvolumes"]
    verbs: ["get", "list", "watch", "create", "delete"]
  - apiGroups: [""]
    resources: ["persistentvolumeclaims"]
    verbs: ["get", "list", "watch", "update"]
  - apiGroups: ["storage.k8s.io"]
    resources: ["storageclasses"]
    verbs: ["get", "list", "watch"]
  - apiGroups: [""]
    resources: ["events"]
    verbs: ["create", "update", "patch"]
  - apiGroups: [""]
    resources: ["endpoints"]
    verbs: ["get", "list", "watch", "create", "update", "patch"]
  - apiGroups: [""]
    resources: ["services"]
    resourceNames: ["kube-dns"]
    verbs: ["list", "get"]
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: rbd-provisioner
subjects:
  - kind: ServiceAccount
    name: rbd-provisioner
    namespace: kube-system
roleRef:
  kind: ClusterRole
  name: rbd-provisioner
  apiGroup: rbac.authorization.k8s.io

---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: rbd-provisioner
  namespace: kube-system
rules:
- apiGroups: [""]
  resources: ["secrets"]
  verbs: ["get"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: rbd-provisioner
  namespace: kube-system
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: Role
  name: rbd-provisioner
subjects:
- kind: ServiceAccount
  name: rbd-provisioner
  namespace: kube-system

---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: rbd-provisioner
  namespace: kube-system
spec:
  replicas: 1
  strategy:
    type: Recreate
  template:
    metadata:
      labels:
        app: rbd-provisioner
    spec:
      containers:
      - name: rbd-provisioner
        image: "quay.io/external_storage/rbd-provisioner:v2.0.0-k8s1.11"
        env:
        - name: PROVISIONER_NAME
          value: ceph.com/rbd
      serviceAccount: rbd-provisioner
EOF
kubectl apply -f external-storage-rbd-provisioner.yaml

# 查看狀態 等待running以後 再進行後續的操做
kubectl get pod -n kube-system
複製代碼

配置 storageclass

# 在k8s集羣中全部節點安裝 ceph-common
# 須要使用kubelet使用rdb命令map附加rbd建立的image
yum install -y ceph-common

# 建立 osd pool 在ceph的mon或者admin節點
ceph osd pool create kube 4096
ceph osd pool ls

# 建立k8s訪問ceph的用戶 在ceph的mon或者admin節點
ceph auth get-or-create client.kube mon 'allow r' osd 'allow class-read object_prefix rbd_children, allow rwx pool=kube' -o ceph.client.kube.keyring

# 查看key 在ceph的mon或者admin節點
ceph auth get-key client.admin
ceph auth get-key client.kube

# 建立 admin secret
# CEPH_ADMIN_SECRET 替換爲 client.admin 獲取到的key
export CEPH_ADMIN_SECRET='AQBBAnRbSiSOFxAAEZXNMzYV6hsceccYLhzdWw=='
kubectl create secret generic ceph-secret --type="kubernetes.io/rbd" \
--from-literal=key=$CEPH_ADMIN_SECRET \
--namespace=kube-system

# 在 default 命名空間建立pvc用於訪問ceph的 secret
# CEPH_KUBE_SECRET 替換爲 client.kube 獲取到的key
export CEPH_KUBE_SECRET='AQBZK3VbTN/QOBAAIYi6CRLQcVevW5HM8lunOg=='
kubectl create secret generic ceph-user-secret --type="kubernetes.io/rbd" \
--from-literal=key=$CEPH_KUBE_SECRET \
--namespace=default

# 查看 secret
kubectl get secret ceph-user-secret -o yaml
kubectl get secret ceph-secret -n kube-system -o yaml

# 配置 StorageClass
# 若是使用kubeadm建立的集羣 provisioner 使用以下方式
# provisioner: ceph.com/rbd
cat >storageclass-ceph-rdb.yaml<<EOF
kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: dynamic-ceph-rdb
provisioner: ceph.com/rbd
# provisioner: kubernetes.io/rbd
parameters:
  monitors: 11.11.11.111:6789,11.11.11.112:6789,11.11.11.113:6789
  adminId: admin
  adminSecretName: ceph-secret
  adminSecretNamespace: kube-system
  pool: kube
  userId: kube
  userSecretName: ceph-user-secret
  fsType: ext4
  imageFormat: "2"
  imageFeatures: "layering"
EOF

# 建立
kubectl apply -f storageclass-ceph-rdb.yaml

# 查看
kubectl get sc
複製代碼

測試使用

# 建立pvc測試
cat >ceph-rdb-pvc-test.yaml<<EOF
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: ceph-rdb-claim
spec:
  accessModes:     
    - ReadWriteOnce
  storageClassName: dynamic-ceph-rdb
  resources:
    requests:
      storage: 2Gi
EOF
kubectl apply -f ceph-rdb-pvc-test.yaml
 
# 查看
kubectl get pvc
kubectl get pv
 
# 建立 nginx pod 掛載測試
cat >nginx-pod.yaml<<EOF
apiVersion: v1
kind: Pod
metadata:
  name: nginx-pod1
  labels:
    name: nginx-pod1
spec:
  containers:
  - name: nginx-pod1
    image: nginx:alpine
    ports:
    - name: web
      containerPort: 80
    volumeMounts:
    - name: ceph-rdb
      mountPath: /usr/share/nginx/html
  volumes:
  - name: ceph-rdb
    persistentVolumeClaim:
      claimName: ceph-rdb-claim
EOF
kubectl apply -f nginx-pod.yaml
 
# 查看
kubectl get pods -o wide
 
# 修改文件內容
kubectl exec -ti nginx-pod1 -- /bin/sh -c 'echo Hello World from Ceph RBD!!! > /usr/share/nginx/html/index.html'
 
# 訪問測試
POD_ID=$(kubectl get pods -o wide | grep nginx-pod1 | awk '{print $(NF-1)}')
curl http://$POD_ID

# 清理
kubectl delete -f nginx-pod.yaml
kubectl delete -f ceph-rdb-pvc-test.yaml
複製代碼

使用 CephFS

linux內核須要4.10+，不然會出現沒法正常使用的問題，詳細issue信息 github.com/kubernetes-… centos7升級內核

在ceph集羣建立CephFS

# 以下操做在ceph的mon或者admin節點

# CephFS須要使用兩個Pool來分別存儲數據和元數據
ceph osd pool create fs_data 128
ceph osd pool create fs_metadata 128
ceph osd lspools

# 建立一個CephFS
ceph fs new cephfs fs_metadata fs_data

# 查看
ceph fs ls
複製代碼

部署cephfs-provisioner

# 官方沒有cephfs動態卷支持
# 使用社區提供的cephfs-provisioner
cat >external-storage-cephfs-provisioner.yaml<<EOF
apiVersion: v1
kind: ServiceAccount
metadata:
  name: cephfs-provisioner
  namespace: kube-system
---
kind: ClusterRole
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: cephfs-provisioner
rules:
  - apiGroups: [""]
    resources: ["persistentvolumes"]
    verbs: ["get", "list", "watch", "create", "delete"]
  - apiGroups: [""]
    resources: ["persistentvolumeclaims"]
    verbs: ["get", "list", "watch", "update"]
  - apiGroups: ["storage.k8s.io"]
    resources: ["storageclasses"]
    verbs: ["get", "list", "watch"]
  - apiGroups: [""]
    resources: ["events"]
    verbs: ["create", "update", "patch"]
  - apiGroups: [""]
    resources: ["endpoints"]
    verbs: ["get", "list", "watch", "create", "update", "patch"]
  - apiGroups: [""]
    resources: ["secrets"]
    verbs: ["create", "get", "delete"]
---
kind: ClusterRoleBinding
apiVersion: rbac.authorization.k8s.io/v1
metadata:
  name: cephfs-provisioner
subjects:
  - kind: ServiceAccount
    name: cephfs-provisioner
    namespace: kube-system
roleRef:
  kind: ClusterRole
  name: cephfs-provisioner
  apiGroup: rbac.authorization.k8s.io

---
apiVersion: rbac.authorization.k8s.io/v1
kind: Role
metadata:
  name: cephfs-provisioner
  namespace: kube-system
rules:
  - apiGroups: [""]
    resources: ["secrets"]
    verbs: ["create", "get", "delete"]
---
apiVersion: rbac.authorization.k8s.io/v1
kind: RoleBinding
metadata:
  name: cephfs-provisioner
  namespace: kube-system
roleRef:
  apiGroup: rbac.authorization.k8s.io
  kind: Role
  name: cephfs-provisioner
subjects:
- kind: ServiceAccount
  name: cephfs-provisioner
  namespace: kube-system

---
apiVersion: extensions/v1beta1
kind: Deployment
metadata:
  name: cephfs-provisioner
  namespace: kube-system
spec:
  replicas: 1
  strategy:
    type: Recreate
  template:
    metadata:
      labels:
        app: cephfs-provisioner
    spec:
      containers:
      - name: cephfs-provisioner
        image: "quay.io/external_storage/cephfs-provisioner:v2.0.0-k8s1.11"
        env:
        - name: PROVISIONER_NAME
          value: ceph.com/cephfs
        command:
        - "/usr/local/bin/cephfs-provisioner"
        args:
        - "-id=cephfs-provisioner-1"
      serviceAccount: cephfs-provisioner
EOF
kubectl apply -f external-storage-cephfs-provisioner.yaml

# 查看狀態 等待running以後 再進行後續的操做
kubectl get pod -n kube-system
複製代碼

#####配置 storageclass

# 查看key 在ceph的mon或者admin節點
ceph auth get-key client.admin

# 建立 admin secret
# CEPH_ADMIN_SECRET 替換爲 client.admin 獲取到的key
# 若是在測試 ceph rbd 方式已經添加 能夠略過此步驟
export CEPH_ADMIN_SECRET='AQBBAnRbSiSOFxAAEZXNMzYV6hsceccYLhzdWw=='
kubectl create secret generic ceph-secret --type="kubernetes.io/rbd" \
--from-literal=key=$CEPH_ADMIN_SECRET \
--namespace=kube-system

# 查看 secret
kubectl get secret ceph-secret -n kube-system -o yaml

# 配置 StorageClass
cat >storageclass-cephfs.yaml<<EOF
kind: StorageClass
apiVersion: storage.k8s.io/v1
metadata:
  name: dynamic-cephfs
provisioner: ceph.com/cephfs
parameters:
    monitors: 11.11.11.111:6789,11.11.11.112:6789,11.11.11.113:6789
    adminId: admin
    adminSecretName: ceph-secret
    adminSecretNamespace: "kube-system"
    claimRoot: /volumes/kubernetes
EOF

# 建立
kubectl apply -f storageclass-cephfs.yaml

# 查看
kubectl get sc
複製代碼

測試使用

# 建立pvc測試
cat >cephfs-pvc-test.yaml<<EOF
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: cephfs-claim
spec:
  accessModes:     
    - ReadWriteOnce
  storageClassName: dynamic-cephfs
  resources:
    requests:
      storage: 2Gi
EOF
kubectl apply -f cephfs-pvc-test.yaml
 
# 查看
kubectl get pvc
kubectl get pv
 
# 建立 nginx pod 掛載測試
cat >nginx-pod.yaml<<EOF
apiVersion: v1
kind: Pod
metadata:
  name: nginx-pod1
  labels:
    name: nginx-pod1
spec:
  containers:
  - name: nginx-pod1
    image: nginx:alpine
    ports:
    - name: web
      containerPort: 80
    volumeMounts:
    - name: cephfs
      mountPath: /usr/share/nginx/html
  volumes:
  - name: cephfs
    persistentVolumeClaim:
      claimName: cephfs-claim
EOF
kubectl apply -f nginx-pod.yaml
 
# 查看
kubectl get pods -o wide
 
# 修改文件內容
kubectl exec -ti nginx-pod1 -- /bin/sh -c 'echo Hello World from CephFS!!! > /usr/share/nginx/html/index.html'
 
# 訪問測試
POD_ID=$(kubectl get pods -o wide | grep nginx-pod1 | awk '{print $(NF-1)}')
curl http://$POD_ID

# 清理
kubectl delete -f nginx-pod.yaml
kubectl delete -f cephfs-pvc-test.yaml
複製代碼

參考文檔

• kubernetes.io/docs/concep…
• docs.openshift.com/container-p…
• ieevee.com/tech/2018/0…
• github.com/kubernetes-…
• github.com/kubernetes-…
• github.com/heketi/heke…
• github.com/gluster/glu…
• github.com/gluster/glu…
• jimmysong.io/kubernetes-…
• kubernetes.io/docs/concep…
• docs.openshift.com/enterprise/…