helm安装mysql8.0集群模式

1.安装operator：MysqlCluster

https://github.com/radondb/radondb-mysql-kubernetes.git

#设置repo
helm repo add radondb https://radondb.github.io/radondb-mysql-kubernetes/
#拉取operator，可进行修改
helm pull radondb/mysql-operator --version=3.0
#安装operator
helm install demo ../mysql-operator

#已将修改后的operator上传到harbor
helm repo add  gwhelm https://harbor.goldwind.com/chartrepo/helm-repo  --ca-file xx.crt --cert-file xx.crt --key-file xx.key  --username 'xx' --password 'xx'
#下载安装helm
helm pull gwhelm/mysql-cluster

2.使用operator安装mysql8.0集群

版本信息：https://github.com/radondb/radondb-mysql-kubernetes/releases

创建mysql8集群yaml: https://github.com/radondb/radondb-mysql-kubernetes/releases/download/v3.0.0/mysql_v1alpha1_mysqlcluster_mysql8.yaml

apiVersion: mysql.radondb.com/v1alpha1
kind: MysqlCluster
metadata:
  name: mysql-cluster
spec:
  replicas: 3
  mysqlVersion: "8.0"
  # the backupSecretName specify the secret file name which store S3 information,
  # if you want S3 backup or restore, please create backup_secret.yaml, uncomment below and fill secret name:
  # backupSecretName: 
  
  # if you want create mysqlcluster from S3, uncomment and fill the directory in S3 bucket below:
  # restoreFrom: 
  
  mysqlOpts:
    image: harbor.goldwind.com/library/percona/percona-server:8.0.25
    user: project_user
    password: wHrr*48+M
    database: project_database
    ## tokudb is not supported in mysql8.
    # initTokuDB: false

    # A simple map between string and string.
    # Such as:
    #    mysqlConf:
    #      expire_logs_days: "7"
    mysqlConf: {}

    resources:
      requests:
        cpu: 256m
        memory: 256Mi
      limits:
        cpu: 1000m
        memory: 1Gi

  xenonOpts:
    image: harbor.goldwind.com/library/radondb/xenon:v2.3.0
    admitDefeatHearbeatCount: 5
    electionTimeout: 10000

    resources:
      requests:
        cpu: 50m
        memory: 128Mi
      limits:
        cpu: 100m
        memory: 256Mi

  metricsOpts:
    enabled: false
    image: harbor.goldwind.com/library/prom/mysqld-exporter:v0.12.1

    resources:
      requests:
        cpu: 10m
        memory: 32Mi
      limits:
        cpu: 100m
        memory: 128Mi

  podPolicy:
    imagePullPolicy: IfNotPresent
    sidecarImage: harbor.goldwind.com/library/radondb/mysql80-sidecar:v2.3.0
    busyboxImage: harbor.goldwind.com/library/busybox:1.32

    slowLogTail: false
    auditLogTail: false

    labels: {}
    annotations: {}
    affinity: 
      podAntiAffinity:
        requiredDuringSchedulingIgnoredDuringExecution:
        - labelSelector:
            matchExpressions:
            - key: app.kubernetes.io/name
              operator: In
              values:
              - mysql
          topologyKey: "kubernetes.io/hostname"
    priorityClassName: ""
    tolerations: []
    schedulerName: ""
    # extraResources defines quotas for containers other than mysql or xenon.
    extraResources:
      requests:
        cpu: 10m
        memory: 32Mi
      limits:
        cpu: 100m
        memory: 125Mi

  persistence:
    enabled: true
    accessModes:
    - ReadWriteOnce
    storageClass: "managed-nfs-storage"
    size: 20Gi

可以在harbor的右上角的kubectl shell中vim写mysql.yaml文件，然后kubectl apply -f mysql.yaml

常见问题：

1）在创建完成后，由于/var/lib/mysql的权限问题 mysql-cluster-2会重复失败重启，此时需要在backup的容器中授权chown -R mysql /var/lib/mysql 因为在backup中使用的是root权限，而在mysql和xenon容器中是mysql权限，且backup容器先创建，所以/var/lib/mysql会是root权限，而mysql和xenon容器对其无权限

2）mysql-cluster-2会报错

get.client.error[dial tcp 10.244.0.35:8801: connect: connection refused]

goroutine 1 [running]:

xbase/xlog.(*Log).Panic(0xc0002062a0, 0x7f67a6, 0x2, 0xc0000f5d08, 0x1, 0x1)
  /go/src/github.com/radondb/xenon/src/xbase/xlog/xlog.go:164 +0x157

查看mysql-cluster-2的日志kubectl logs --tail=3000 -f pod/mysql-cluster-mysql-0 -n demo 可以看到报错

2023-07-21T10:52:47.948569+08:00 0 [ERROR] [MY-010544] [Repl] Failed to open the relay log '/var/lib/mysql/mysql-relay-bin.000002' (relay_log_pos 371).
2023-07-21T10:52:47.949938+08:00 0 [ERROR] [MY-011059] [Repl] Could not find target log file mentioned in relay log info in the index file '/var/lib/mysql/mysql-relay-bin.index' during relay log initialization.
2023-07-21T10:52:47.951793+08:00 0 [ERROR] [MY-010426] [Repl] Slave: Failed to initialize the master info structure for channel ''; its record may still be present in 'mysql.slave_master_info' table, consider deleting it.
2023-07-21T10:52:47.953076+08:00 0 [ERROR] [MY-010529] [Repl] Failed to create or recover replication info repositories.
2023-07-21T10:52:48.116508+08:00 0 [System] [MY-010931] [Server] /usr/sbin/mysqld: ready for connections. Version: '8.0.25-15'  socket: '/var/lib/mysql/mysql.sock'  port: 3306  Percona Server (GPL), Release 15, Revision a558ec2.
2023-07-21T10:52:49.996244+08:00 8 [System] [MY-010597] [Repl] 'CHANGE MASTER TO FOR CHANNEL '' executed'. Previous state master_host='', master_port= 3306, master_log_file='', master_log_pos= 4, master_bind=''. New state master_host='mysql-cluster-mysql-0.mysql-cluster-mysql.wfs-prod', master_port= 3306, master_log_file='', master_log_pos= 4, master_bind=''.
2023-07-21T10:52:50.013805+08:00 10 [Warning] [MY-010897] [Repl] Storing MySQL user name or password information in the master info repository is not secure and is therefore not recommended. Please consider using the USER and PASSWORD connection options for START SLAVE; see the 'START SLAVE Syntax' in the MySQL Manual for more information.
2023-07-21T10:52:50.017150+08:00 10 [System] [MY-010562] [Repl] Slave I/O thread for channel '': connected to master 'radondb_repl@mysql-cluster-mysql-0.mysql-cluster-mysql.wfs-prod:3306',replication started in log 'FIRST' at position 4
2023-07-21T10:53:03.460369+08:00 10 [ERROR] [MY-010557] [Repl] Error reading packet from server for channel '': Lost connection to MySQL server during query (server_errno=2013)
2023-07-21T10:53:03.460500+08:00 10 [Warning] [MY-010897] [Repl] Storing MySQL user name or password information in the master info repository is not secure and is therefore not recommended. Please consider using the USER and PASSWORD connection options for START SLAVE; see the 'START SLAVE Syntax' in the MySQL Manual for more information.
2023-07-21T10:53:03.462175+08:00 10 [ERROR] [MY-010584] [Repl] Slave I/O for channel '': error reconnecting to master 'radondb_repl@mysql-cluster-mysql-0.mysql-cluster-mysql.wfs-prod:3306' - retry-time: 60 retries: 1 message: Can't connect to MySQL server on 'mysql-cluster-mysql-0.mysql-cluster-mysql.wfs-prod:3306' (111), Error_code: MY-002003