Kubernetes StatefulSet 实战3节点MySQL集群部署与数据持久化验证在云原生架构中数据库集群的部署一直是DevOps工程师面临的核心挑战之一。传统虚拟机环境下的数据库部署方式在Kubernetes集群中往往水土不服而StatefulSet作为Kubernetes中管理有状态应用的核心控制器为这一难题提供了优雅的解决方案。本文将深入探讨如何利用StatefulSet部署一个生产可用的3节点MySQL集群并验证其数据持久化能力。1. 环境准备与基础概念在开始部署之前我们需要明确几个关键概念。StatefulSet与常见的Deployment控制器不同它为每个Pod提供稳定的标识符和持久化存储这对于数据库这类有状态应用至关重要。StatefulSet的核心特性稳定的网络标识每个Pod拥有固定的DNS名称格式为pod-name.service-name有序部署与扩展Pod按顺序创建从0到N-1和终止从N-1到0持久化存储通过VolumeClaimTemplate为每个Pod动态创建PVC1.1 前置条件检查确保您的Kubernetes集群满足以下要求# 检查Kubernetes版本 kubectl version --short # 检查StorageClass可用性 kubectl get storageclass如果您的集群没有配置动态存储供应需要预先创建PV。以下是创建3个1Gi PV的示例apiVersion: v1 kind: PersistentVolume metadata: name: mysql-pv-0 spec: capacity: storage: 1Gi accessModes: - ReadWriteOnce hostPath: path: /data/mysql-0 --- apiVersion: v1 kind: PersistentVolume metadata: name: mysql-pv-1 spec: capacity: storage: 1Gi accessModes: - ReadWriteOnce hostPath: path: /data/mysql-1 --- apiVersion: v1 kind: PersistentVolume metadata: name: mysql-pv-2 spec: capacity: storage: 1Gi accessModes: - ReadWriteOnce hostPath: path: /data/mysql-22. MySQL集群部署架构我们的3节点MySQL集群将采用主从复制架构其中mysql-0作为主节点mysql-1和mysql-2作为从节点使用Headless Service提供稳定的DNS解析每个Pod拥有独立的持久化存储2.1 创建命名空间和Secret首先为MySQL集群创建独立的命名空间apiVersion: v1 kind: Namespace metadata: name: mysql-cluster然后创建包含MySQL root密码的SecretapiVersion: v1 kind: Secret metadata: name: mysql-secret namespace: mysql-cluster type: Opaque data: password: MTIzNDU2 # echo -n 123456 | base642.2 部署Headless ServiceHeadless Service是StatefulSet的核心组件它为每个Pod提供稳定的DNS记录apiVersion: v1 kind: Service metadata: name: mysql namespace: mysql-cluster labels: app: mysql spec: ports: - name: mysql port: 3306 clusterIP: None selector: app: mysql3. StatefulSet配置详解以下是完整的MySQL StatefulSet配置我们将分段解析关键部分3.1 基础配置apiVersion: apps/v1 kind: StatefulSet metadata: name: mysql namespace: mysql-cluster spec: serviceName: mysql replicas: 3 selector: matchLabels: app: mysql3.2 Pod模板配置template: metadata: labels: app: mysql spec: initContainers: - name: init-mysql image: mysql:5.7 command: - bash - -c - | set -ex # 根据Pod序号生成server-id [[ hostname ~ -([0-9])$ ]] || exit 1 ordinal${BASH_REMATCH[1]} echo [mysqld] /mnt/conf.d/server-id.cnf echo server-id$((100 $ordinal)) /mnt/conf.d/server-id.cnf # 主节点配置binlog if [[ $ordinal -eq 0 ]]; then echo log-binmysql-bin /mnt/conf.d/master.cnf else echo log-slave-updates1 /mnt/conf.d/slave.cnf fi volumeMounts: - name: conf mountPath: /mnt/conf.d - name: config-map mountPath: /mnt/config-map3.3 数据持久化配置volumeClaimTemplates: - metadata: name: data spec: accessModes: [ReadWriteOnce] resources: requests: storage: 1Gi完整配置还包含容器定义、健康检查等部分限于篇幅不在此展示。应用配置kubectl apply -f mysql-statefulset.yaml4. 集群初始化与配置验证4.1 观察Pod创建过程StatefulSet会按顺序创建Podkubectl get pods -n mysql-cluster -w输出应显示三个Pod依次进入Running状态NAME READY STATUS RESTARTS AGE mysql-0 1/1 Running 0 2m mysql-1 1/1 Running 0 1m mysql-2 1/1 Running 0 30s4.2 验证DNS解析kubectl run -it --rm --imagemysql:5.7 --restartNever mysql-client -- bash在容器内执行nslookup mysql-0.mysql.mysql-cluster.svc.cluster.local nslookup mysql-1.mysql.mysql-cluster.svc.cluster.local nslookup mysql-2.mysql.mysql-cluster.svc.cluster.local5. 数据持久化验证5.1 写入测试数据kubectl exec -it mysql-0 -n mysql-cluster -- mysql -uroot -p123456在MySQL命令行执行CREATE DATABASE test; USE test; CREATE TABLE t1 (id INT PRIMARY KEY, name VARCHAR(20)); INSERT INTO t1 VALUES (1, data_from_mysql0);5.2 模拟Pod故障删除mysql-0 Pod验证数据持久性kubectl delete pod mysql-0 -n mysql-cluster等待新Pod创建完成后验证数据kubectl exec -it mysql-0 -n mysql-cluster -- mysql -uroot -p123456 -e SELECT * FROM test.t15.3 存储卷验证检查PVC状态kubectl get pvc -n mysql-cluster输出应显示三个PVC都处于Bound状态NAME STATUS VOLUME CAPACITY ACCESS MODES>kubectl scale statefulset mysql -n mysql-cluster --replicas56.2 滚动更新策略StatefulSet支持多种更新策略以下是配置示例spec: updateStrategy: type: RollingUpdate rollingUpdate: partition: 2 # 只有序号2的Pod会被更新7. 生产环境优化建议7.1 资源配置建议resources: requests: memory: 2Gi cpu: 500m limits: memory: 4Gi cpu: 27.2 监控与告警建议配置以下监控指标MySQL连接数查询吞吐量复制延迟存储空间使用率7.3 备份策略定期备份PVC数据可以使用Velero等工具velero backup create mysql-backup \ --include-namespaces mysql-cluster \ --include-resources persistentvolumeclaims,persistentvolumes