隐私政策
留言板
金色传说
kubernetes
terraform
云生原
helm
代码编程
Java
Python
Shell
DevOps
Ansible
Gitlab
Jenkins
运维
老司机
Linux 杂锦
Nginx
数据库
elasticsearch
监控
上帝视角
DJI FPV
DJI mini 3 pro
关于本站共计 9555 个字符，预计需要花费 24 分钟才能阅读完成。
StatefulSet是一种有状态的资源集合，管理所有有状态的服务，常用于Mysql/MongoDB集群等。StatefulSet所管理的Pod有如下特点：
- 具有固定网络标识的Pod名称和hostname，(statefulset名称)−(序号)，序号：[0，副本数-1)
- Pod顺序启停
- 稳定的存储，通过VolumeClaimTemplate为每个Pod创建一一对应的PV，删除或减少副本数，不会移除相关pv。
- 其对应的Service为headless service（解析该名称将会放回对应所有的Pod的Endpoint列表）
创建StatefulSet 测试资源
# 创建一个demo
[root@node1 pratic]# cat >statusful-nginx.yaml<<EOF
apiVersion: v1
kind: Service
metadata:
name: nginx
labels:
app: nginx
spec:
ports:
- port: 80
name: web
clusterIP: None
selector:
app: nginx
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: web
spec:
serviceName: "nginx"
replicas: 5
selector:
matchLabels:
app: nginx
template:
metadata:
labels:
app: nginx
spec:
containers:
- name: nginx
image: nginx:latest
ports:
- containerPort: 80
name: web
volumeMounts:
- name: www
mountPath: /usr/share/nginx/html
volumeClaimTemplates:
- metadata:
name: www
annotations:
volume.beta.kubernetes.io/storage-class: "nfsserver"
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 100Mi
EOFPod启停有序性
观察StatusfulSet Pod创建顺序
web-0 pod处于Running和Ready状态后web-1 Pod才会被启动，这就是有状态集的启动有序性
[root@node1 pratic]# kubectl apply -f statusful-nginx.yaml
[root@node1 pratic]# kubectl get pods -w
NAME READY STATUS RESTARTS AGE
web-0 1/1 Running 0 32s
web-1 1/1 Running 0 18s
web-2 0/1 ContainerCreating 0 8s
web-2 1/1 Running 0 16s
web-3 0/1 Pending 0 0s
web-3 0/1 Pending 0 0s
web-3 0/1 Pending 0 2s
web-3 0/1 ContainerCreating 0 2s
web-3 0/1 ContainerCreating 0 3s
web-3 1/1 Running 0 12s
web-4 0/1 Pending 0 0s
web-4 0/1 Pending 0 0s
web-4 0/1 Pending 0 2s
web-4 0/1 ContainerCreating 0 2s
web-4 0/1 ContainerCreating 0 3s
web-4 1/1 Running 0 9sPod网络标志固定
测试每个pod的hostname
[root@node1 pratic]# for i in {0..4}; do kubectl exec "web-$i" -- sh -c 'hostname'; done
web-0
web-1
web-2
web-3
web-4测试Pod的DNS地址
# 创建一个静态pod busybox
[root@node1 pratic]# kubectl run -i --tty --image busybox:1.28 dns-test --restart=Never --rm
If you don't see a command prompt, try pressing enter.
/ # nslookup web-0.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-0.nginx
Address 1: 10.100.166.172 web-0.nginx.default.svc.cluster.local
/ # nslookup web-1.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-1.nginx
Address 1: 10.100.166.173 web-1.nginx.default.svc.cluster.local
/ # nslookup web-2.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-2.nginx
Address 1: 10.100.166.174 web-2.nginx.default.svc.cluster.local测试固定的网络标识
1.先删除创建的Pod
# 先删除创建的demo pod，再删除的同时开启另一个终端查看demo pod资源创建情况
[root@node1 pratic]# kubectl delete pod -l app=nginx
pod "web-0" deleted
pod "web-1" deleted
pod "web-2" deleted
pod "web-3" deleted
pod "web-4" deleted
# 另一终端查看资源创建情况
[root@node1 pratic]# kubectl get pod -w -l app=nginx
NAME web-0 1/1 Terminating 0 5m2s
web-1 1/1 Terminating 0 4m52s
web-2 1/1 Terminating 0 4m45s
web-3 1/1 Terminating 0 4m37s
web-4 1/1 Terminating 0 4m27s
web-1 0/1 Terminating 0 4m54s
web-4 0/1 Terminating 0 4m29s
web-2 0/1 Terminating 0 4m47s
web-3 0/1 Terminating 0 4m39s
web-0 0/1 Terminating 0 5m4s
web-3 0/1 Terminating 0 4m45s
web-3 0/1 Terminating 0 4m45s
web-0 0/1 Terminating 0 5m10s
web-0 0/1 Terminating 0 5m10s
web-1 0/1 Terminating 0 5m
web-0 0/1 Pending 0 0s
web-1 0/1 Terminating 0 5m
web-0 0/1 Pending 0 0s
web-2 0/1 Terminating 0 4m53s
web-2 0/1 Terminating 0 4m53s
web-4 0/1 Terminating 0 4m35s
web-4 0/1 Terminating 0 4m35s
web-0 0/1 ContainerCreating 0 1s
web-0 0/1 ContainerCreating 0 2s
web-0 1/1 Running 0 8s
web-1 0/1 Pending 0 0s
web-1 0/1 Pending 0 0s
web-1 0/1 ContainerCreating 0 0s
web-1 0/1 ContainerCreating 0 1s
web-1 1/1 Running 0 9s
web-2 0/1 Pending 0 0s
web-2 0/1 Pending 0 0s
web-2 0/1 ContainerCreating 0 0s
web-2 0/1 ContainerCreating 0 2s
web-2 1/1 Running 0 11s
web-3 0/1 Pending 0 0s
web-3 0/1 Pending 0 0s
web-3 0/1 ContainerCreating 0 0s
web-3 0/1 ContainerCreating 0 1s
web-3 1/1 Running 0 7s
web-4 0/1 Pending 0 0s
web-4 0/1 Pending 0 0s
web-4 0/1 ContainerCreating 0 0s
web-4 0/1 ContainerCreating 0 1s
web-4 1/1 Running 0 9s2.测试此时重建的pod的网络标志
# 查看重建pod后的hostname
[root@node1 pratic]# for i in {0..4}; do kubectl exec web-$i -- sh -c 'hostname'; done
web-0
web-1
web-2
web-3
web-4
查看重建pod后的dns记录
/ # nslookup web-0.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-0.nginx
Address 1: 10.100.166.188 10-100-166-188.nginx-service.default.svc.cluster.local
/ # nslookup web-1.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-1.nginx
Address 1: 10.100.166.189 web-1.nginx.default.svc.cluster.local
/ # nslookup web-2.nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: web-2.nginx
Address 1: 10.100.166.190 10-100-166-190.nginx-service.default.svc.cluster.local
# 获取到
/ # nslookup nginx
Server: 10.96.0.10
Address 1: 10.96.0.10 kube-dns.kube-system.svc.cluster.local
Name: nginx
Address 1: 10.100.166.190 web-2.nginx.default.svc.cluster.local
Address 2: 10.100.166.189 10-100-166-189.nginx-service.default.svc.cluster.local
Address 3: 10.100.166.188 10-100-166-188.nginx-service.default.svc.cluster.local
Address 4: 10.100.166.134 10-100-166-134.nginx-service.default.svc.cluster.local
Address 5: 10.100.166.136 10-100-166-136.nginx-service.default.svc.cluster.local从上面测试可以看出Pod 的序号、主机名、SRV 条目和记录名称没有改变，但和 Pod 相关联的 IP 地址可能发生了改变。所以在应用中使用StatefulSet时应该使用pod的SRV记录。
稳定的存储
查看测试资源创建的pvc/pv，发现其分别为每个pod创建了一个pvc，且是一一根据序号对应
# 查看pvc
[root@node1 pratic]# kubectl get pvc -l app=nginx
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
www-web-0 Bound pvc-a08aa8a7-cd1d-4584-b55b-80cd1b217e66 100Mi RWO nfsserver 33m
www-web-1 Bound pvc-5707ad69-9e7c-4423-9144-e3e2f000b9ed 100Mi RWO nfsserver 33m
www-web-2 Bound pvc-975520a6-de1f-4e26-b7d8-65e3353866ab 100Mi RWO nfsserver 33m
www-web-3 Bound pvc-4e7b5b31-a334-4503-8b82-935d2a561cc0 100Mi RWO nfsserver 33m
www-web-4 Bound pvc-39787c7a-2ae3-44ca-8821-d2bad2a4a75c 100Mi RWO nfsserver 33m
# 查看pvc
[root@node1 pratic]# kubectl get pv | grep www-web
pvc-39787c7a-2ae3-44ca-8821-d2bad2a4a75c 100Mi RWO Delete Bound default/www-web-4 nfsserver 34m
pvc-4e7b5b31-a334-4503-8b82-935d2a561cc0 100Mi RWO Delete Bound default/www-web-3 nfsserver 34m
pvc-5707ad69-9e7c-4423-9144-e3e2f000b9ed 100Mi RWO Delete Bound default/www-web-1 nfsserver 35m
pvc-975520a6-de1f-4e26-b7d8-65e3353866ab 100Mi RWO Delete Bound default/www-web-2 nfsserver 35m
pvc-a08aa8a7-cd1d-4584-b55b-80cd1b217e66 100Mi RWO Delete Bound default/www-web-0 nfsserver 35m测试pvc和pod的对应关系
1.将每个Pod hostname写入各自PV中
[root@node1 pratic]# for i in {0..4}; do kubectl exec "web-$i" -- sh -c 'echo "$(hostname)" > /usr/share/nginx/html/index.html'; done
[root@node1 pratic]# for i in {0..4}; do kubectl exec -i -t "web-$i" -- curl http://localhost/; done
web-0
web-1
web-2
web-3
web-42.删除所有Pod
[root@node1 pratic]# kubectl delete pod -l app=nginx
pod "web-0" deleted
pod "web-1" deleted
pod "web-2" deleted
pod "web-3" deleted
pod "web-4" deleted3.查看pv/pvc
# 五个 PersistentVolumeClaims 和五个 PersistentVolumes 仍然存在
[root@node1 pratic]# kubectl get pvc -l app=nginx
NAME STATUS VOLUME CAPACITY ACCESS MODES STORAGECLASS AGE
www-web-0 Bound pvc-a08aa8a7-cd1d-4584-b55b-80cd1b217e66 100Mi RWO nfsserver 33m
www-web-1 Bound pvc-5707ad69-9e7c-4423-9144-e3e2f000b9ed 100Mi RWO nfsserver 33m
www-web-2 Bound pvc-975520a6-de1f-4e26-b7d8-65e3353866ab 100Mi RWO nfsserver 33m
www-web-3 Bound pvc-4e7b5b31-a334-4503-8b82-935d2a561cc0 100Mi RWO nfsserver 33m
www-web-4 Bound pvc-39787c7a-2ae3-44ca-8821-d2bad2a4a75c 100Mi RWO nfsserver 33m
[root@node1 pratic]# kubectl get pv | grep www-web
pvc-39787c7a-2ae3-44ca-8821-d2bad2a4a75c 100Mi RWO Delete Bound default/www-web-4 nfsserver 34m
pvc-4e7b5b31-a334-4503-8b82-935d2a561cc0 100Mi RWO Delete Bound default/www-web-3 nfsserver 34m
pvc-5707ad69-9e7c-4423-9144-e3e2f000b9ed 100Mi RWO Delete Bound default/www-web-1 nfsserver 35m
pvc-975520a6-de1f-4e26-b7d8-65e3353866ab 100Mi RWO Delete Bound default/www-web-2 nfsserver 35m
pvc-a08aa8a7-cd1d-4584-b55b-80cd1b217e66 100Mi RWO Delete Bound default/www-web-0 nfsserver 35m4.验证每个Pod和PV bond关系，虽然所有Pod重新创建和调度，但是他们PersistentVolumeClaim 相关联的 PersistentVolume 被重新挂载到了各自的 volumeMount 上。
[root@node1 pratic]# for i in {0..4}; do kubectl exec -i -t "web-$i" -- curl http://localhost/; done
web-0
web-1
web-2
web-3
web-4StatefulSet扩缩容
扩容
[root@node1 pratic]# kubectl scale sts web --replicas=8
statefulset.apps/web scaled
# 安顺序扩容
[root@node1 pratic]# kubectl get pod -w -l app=nginx
......
web-5 0/1 Pending 0 0s
web-5 0/1 Pending 0 0s
web-5 0/1 Pending 0 1s
web-5 0/1 ContainerCreating 0 1s
web-5 0/1 ContainerCreating 0 2s
web-5 1/1 Running 0 8s
web-6 0/1 Pending 0 0s
web-6 0/1 Pending 0 0s
web-6 0/1 Pending 0 2s
web-6 0/1 ContainerCreating 0 2s
web-6 0/1 ContainerCreating 0 3s
web-6 1/1 Running 0 13s
web-7 0/1 Pending 0 0s
web-7 0/1 Pending 0 0s
web-7 0/1 Pending 0 2s
web-7 0/1 ContainerCreating 0 2s
web-7 0/1 ContainerCreating 0 4s
web-7 1/1 Running 0 11s缩容
[root@node1 pratic]# kubectl scale sts web --replicas=5
statefulset.apps/web scaled
# 按倒序缩容
[root@node1 pratic]# kubectl get pod -w -l app=nginx
......
web-7 1/1 Terminating 0 2m50s
web-7 0/1 Terminating 0 2m52s
web-7 0/1 Terminating 0 2m53s
web-7 0/1 Terminating 0 2m53s
web-6 1/1 Terminating 0 3m6s
web-6 0/1 Terminating 0 3m7s
web-6 0/1 Terminating 0 3m8s
web-6 0/1 Terminating 0 3m8s
web-5 1/1 Terminating 0 3m16s
web-5 0/1 Terminating 0 3m17s
web-5 0/1 Terminating 0 3m25s
web-5 0/1 Terminating 0 3m25spodManagementPolicy
有时候StatefulSet的启停有序性并不是必要的，只需要网络标志和存储固定即可，所有可以使用podManagementPolicy字段控制
[root@node1 pratic]# cat >statusfulset-nginx2.yaml<<EOF
apiVersion: v1
kind: Service
metadata:
name: nginx2
labels:
app: nginx2
spec:
ports:
- port: 80
name: web
clusterIP: None
selector:
app: nginx2
---
apiVersion: apps/v1
kind: StatefulSet
metadata:
name: web2
spec:
serviceName: "nginx2"
podManagementPolicy: "Parallel"
replicas: 4
selector:
matchLabels:
app: nginx2
template:
metadata:
labels:
app: nginx2
spec:
containers:
- name: nginx2
image: nginx:latest
ports:
- containerPort: 80
name: web
volumeMounts:
- name: www2
mountPath: /usr/share/nginx/html
volumeClaimTemplates:
- metadata:
name: www2
annotations:
volume.beta.kubernetes.io/storage-class: "nfsserver"
spec:
accessModes: [ "ReadWriteOnce" ]
resources:
requests:
storage: 100Mi
EOF
测试查看创建Pod的顺序
[root@node1 pratic]# kubectl apply -f statusfulset-nginx2.yaml
service/nginx2 created
statefulset.apps/web2 created
# 不再按顺序等待pod状态为running和ready，而是同时创建
[root@node1 pratic]# kubectl get pod -w -l app=nginx2
NAME READY STATUS RESTARTS AGE
web2-0 0/1 ContainerCreating 0 4s
web2-1 0/1 ContainerCreating 0 4s
web2-2 0/1 ContainerCreating 0 4s
web2-3 0/1 ContainerCreating 0 4s
web2-1 1/1 Running 0 8s
web2-3 1/1 Running 0 13s
web2-0 1/1 Running 0 19s
web2-2 1/1 Running 0 25s
正文完
