k8s-pvc/pv扩容记录
背景
一次聊天过程中，对方提及pvc的扩容，虽然有注意过 storageclass 有个AllowVolumeExpansion的配置（有些csi插件是不支持该配置的，比如local-volume-provisoner），但是没有实际用过，所以还是心里没底，所以抽时间 做了下扩容的操作。这里是基于 ceph-csi 插件，在rbd的使用方式下做的 扩容
原理
需要csi 插件的支持，更改pod的 pvc 申明的存储空间需求，csi插件便会自动对原来的rbd卷 做扩容操作
步骤
##检查当前storage class 支持的特性
wsg@k8snode1v18:~/k8s_deploy/ceph-csi/examples/rbd$ kubectl describe sc csi-rbd-sc
AllowVolumeExpansion: True
MountOptions:
discard
ReclaimPolicy: Delete
VolumeBindingMode: Immediate
Events:
可以看到，AllowVolumeExpansion 为true
创建test pod 并且 申明使用
测试pod 定义如下：
wsg@k8snode1v18:~/k8s_deploy/ceph-csi/examples/rbd$ cat pod-test-pv-expansion.yaml
---
apiVersion: v1
kind: Pod
metadata:
name: pod-test-pvc-expansion
spec:
containers:
- name: nginx-server
image: nginx
volumeMounts:
- name: mypvc
mountPath: /var/lib/www/html
volumes:
- name: mypvc
persistentVolumeClaim:
claimName: rbd-pvc
readOnly: false
---
apiVersion: v1
kind: PersistentVolumeClaim
metadata:
name: rbd-pvc
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 1Gi
storageClassName: csi-rbd-sc
进入测试pod 检查 volume 挂载情况
wsg@k8snode1v18:~/k8s_deploy/ceph-csi/examples/rbd$ kubectl exec -it pod-test-pvc-expansion – /bin/bash
/dev/rbd1 on /var/lib/www/html type ext4 (rw,relatime,stripe=16)
可以看到 声明的 1G rbd 磁盘已经 自动 mount 上了，并且采用 ext4 文件系统
写入测试数据
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod# md5sum before-expansion.txt >md5
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod# md5sum before-expansion.txt
0b4cfe43fc1f774cc0ed294515f626e1 before-expansion.txt
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod# cat md5
0b4cfe43fc1f774cc0ed294515f626e1 before-expansion.txt
root@pod-test-pvc-expansion:/var/lib/www/html# md5sum test.txt |tee md5
798f8e998aa4c359f7ef66549866fb01 test.txt
root@pod-test-pvc-expansion:/var/lib/www/html# cat md5
798f8e998aa4c359f7ef66549866fb01 test.txt
执行扩容（调整 pvc 声明的大小）
spec:
accessModes:
- ReadWriteOnce
resources:
requests:
storage: 10Gi
storageClassName: csi-rbd-sc
volumeMode: Filesystem
volumeName: pvc-abfad8c1-9d71-45fd-834c-60b8109c6b75
wsg@k8snode1v18:~$ kubectl edit pvc rbd-pvc
persistentvolumeclaim/rbd-pvc edited
wsg@k8snode1v18:~$
检查 pod 状态
##扩容中看到 pod 没有重启的记录（扩容前 kubectl exec的 会话 也是一直保持正常，没有中断）
pod-test-pvc-expansion 1/1 Running 0 25m
##进一步检查 pod 的event，发现已经有 成功扩容的记录
wsg@k8snode1v18:~$ kubectl describe pod pod-test-pvc-expansion
Name: pod-test-pvc-expansion
Namespace: default
Priority: 0
Node: k8snode1v18/172.17.73.53
Start Time: Sat, 27 Nov 2021 14:05:46 +0800
Labels:
Annotations: cni.projectcalico.org/podIP: 10.106.176.31/32
Status: Running
IP: 10.106.176.31
IPs:
IP: 10.106.176.31
Containers:
nginx-server:
Container ID: docker://ce3e35282b5ff3dcd62848ab69efb751f592510e67aed9ad3183eaf817f54f21
Image: nginx
Image ID: docker-pullable://nginx@sha256:097c3a0913d7e3a5b01b6c685a60c03632fc7a2b50bc8e35bcaa3691d788226e
Port:
Host Port:
State: Running
Started: Sat, 27 Nov 2021 14:06:35 +0800
Ready: True
Restart Count: 0
Environment:
Mounts:
/var/lib/www/html from mypvc (rw)
/var/run/secrets/kubernetes.io/serviceaccount from default-token-gc5x4 (ro)
Conditions:
Type Status
Initialized True
Ready True
ContainersReady True
PodScheduled True
Volumes:
mypvc:
Type: PersistentVolumeClaim (a reference to a PersistentVolumeClaim in the same namespace)
ClaimName: rbd-pvc
ReadOnly: false
default-token-gc5x4:
Type: Secret (a volume populated by a Secret)
SecretName: default-token-gc5x4
Optional: false
QoS Class: BestEffort
Node-Selectors:
Tolerations: node.kubernetes.io/not-ready:NoExecute for 300s
node.kubernetes.io/unreachable:NoExecute for 300s
Events:
Type Reason Age From Message
---- ------ ---- ---- -------
Warning FailedScheduling default-scheduler persistentvolumeclaim "rbd-pvc" not found
Warning FailedScheduling default-scheduler persistentvolumeclaim "rbd-pvc" not found
Warning FailedScheduling default-scheduler running "VolumeBinding" filter plugin for pod "pod-test-pvc-expansion": pod has unbound immediate PersistentVolumeClaims
Normal Scheduled default-scheduler Successfully assigned default/pod-test-pvc-expansion to k8snode1v18
Normal SuccessfulAttachVolume 17m attachdetach-controller AttachVolume.Attach succeeded for volume "pvc-abfad8c1-9d71-45fd-834c-60b8109c6b75"
Normal Pulling 17m kubelet, k8snode1v18 Pulling image "nginx"
Normal Pulled 16m kubelet, k8snode1v18 Successfully pulled image "nginx"
Normal Created 16m kubelet, k8snode1v18 Created container nginx-server
Normal Started 16m kubelet, k8snode1v18 Started container nginx-server
Normal FileSystemResizeSuccessful 48s kubelet, k8snode1v18 MountVolume.NodeExpandVolume succeeded for volume "pvc-abfad8c1-9d71-45fd-834c-60b8109c6b75"
wsg@k8snode1v18:~$
扩容前后 pvc 的对比
rbd-pvc Bound pvc-abfad8c1-9d71-45fd-834c-60b8109c6b75 1Gi RWO csi-rbd-sc 13m
//扩容后
rbd-pvc Bound pvc-abfad8c1-9d71-45fd-834c-60b8109c6b75 10Gi RWO csi-rbd-sc 19m
wsg@k8snode1v18:~$
可以看到 容量的变化，并且 绑定的 pv编号 也是没有变化的，说明是同一个pv
检查 pod 中的数据
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod# md5sum before-expansion.txt
0b4cfe43fc1f774cc0ed294515f626e1 before-expansion.txt
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod# cat md5
0b4cfe43fc1f774cc0ed294515f626e1 before-expansion.txt
root@pod-test-pvc-expansion:/var/lib/www/html/wsg-inpod#
知识点
pod 可以看到 host mount namespce
root@pod-test-pvc-expansion:/# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 2.5M 1 loop
loop1 7:1 0 99.4M 1 loop
loop2 7:2 0 61.8M 1 loop
loop3 7:3 0 2.5M 1 loop
loop4 7:4 0 61.8M 1 loop
loop5 7:5 0 242.3M 1 loop
loop6 7:6 0 548K 1 loop
loop7 7:7 0 4K 1 loop
loop8 7:8 0 704K 1 loop
loop10 7:10 0 704K 1 loop
loop11 7:11 0 65.1M 1 loop
loop12 7:12 0 140.7M 1 loop
loop13 7:13 0 548K 1 loop
loop14 7:14 0 2.5M 1 loop
loop15 7:15 0 247.9M 1 loop
loop16 7:16 0 219M 1 loop
loop17 7:17 0 2.5M 1 loop
loop18 7:18 0 65.2M 1 loop
loop19 7:19 0 219M 1 loop
loop20 7:20 0 55.5M 1 loop
loop21 7:21 0 140.7M 1 loop
loop22 7:22 0 99.4M 1 loop
loop23 7:23 0 55.5M 1 loop
sda 8:0 0 500G 0 disk
`-sda1 8:1 0 500G 0 part /etc/hosts
rbd0 252:0 0 60G 0 disk
rbd1 252:16 0 1G 0 disk /var/lib/www/html
root@pod-test-pvc-expansion:/#