使用Ceph集群作为Kubernetes的动态分配持久化存储

1

2

# ceph osd pool create k8s 128

pool 'k8s' created

1

# ceph auth add client.k8s mon 'allow rx' osd 'allow rwx pool=k8s'

1

2

3

4

client.k8s

key: AQC3Hm5Zan9LDhAAXZHCdAF39bXcEwdpV6y/cA==

caps: [mon] allow r

caps: [osd] allow rw pool=k8s

1

2

3

4

# rbd create k8s/foo --size 1G --id k8s

# rbd map k8s/foo --id k8s

/dev/rbd0

1

2

# echo "$(ceph auth get-key client.k8s)"|base64

QVFDM0htNVphbjlMRGhBQVhaSENkQUYzOWJYY0V3ZHBWNnkvY0E9PQo=

1

2

3

4

5

6

7

8

apiVersion: v1

kind: Secret

metadata:

name: ceph-secret

namespace: kube-system

type: "kubernetes.io/rbd"

data:

key: "QVFDM0htNVphbjlMRGhBQVhaSENkQUYzOWJYY0V3ZHBWNnkvY0E9PQo="

1

# kubectl create -f ceph-secret.yaml

1

2

# kubectl get secret -n=kube-system|grep ceph

ceph-secret kubernetes.io/rbd 1 1m

1

2

3

4

5

6

7

8

9

10

11

12

13

apiVersion: storage.k8s.io/v1beta1

kind: StorageClass

metadata:

name: fast

provisioner: kubernetes.io/rbd

parameters:

monitors: 172.30.30.215:6789,172.30.30.217:6789,172.30.30.219:6789

adminId: k8s

adminSecretName: ceph-secret

adminSecretNamespace: kube-system

pool: k8s

userId: k8s

userSecretName: ceph-secret

1

# kubectl create -f ceph-rbd-storageclass.yaml

1

2

3

# kubectl get storageclass

NAME TYPE

fast kubernetes.io/rbd

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

{

"kind": "PersistentVolumeClaim",

"apiVersion": "v1",

"metadata": {

"name": "claim1",

"annotations": {

"volume.beta.kubernetes.io/storage-class": "fast"

}

},

"spec": {

"accessModes": [

"ReadWriteOnce"

],

"resources": {

"requests": {

"storage": "3Gi"

}

}

}

}

1

2

3

4

5

6

# kubectl create -f ceph-pvc.json

# kubectl get pvc

NAME STATUS VOLUME CAPACITY ACCESSMODES AGE

claim1 Bound pvc-28b66dcb-6c82-11e7-94da-02672b869d7f 3Gi RWO 11m