Linux mysql 双机集群配置教程

Linuxmysql双机集群配置教程2011-2-1220:17:00来源：Sql一、介绍这篇文档旨在介绍如何安装配置基于2台服务器的MySQL集群。并且实现任意一台服务器出现问题或宕机时MySql集群依然能够继续运行。安装环境及软件包：2台PCmysql-6.0.0-alpha.tar.gzCentos5SerA:0SerB:二、在SerA和SerB上安装MySQL以下步骤需要在SerA和SerB上各做一次mvmysql-6.0.0-alpha.tar.gz/tmp/packagecd/tmp/packagegroupaddmysqluseradd-gmysqlmysqltar-zxvfmysql-6.0.0-alpha.tar.gzrm-fmysql-6.0.0-alpha.tar.gzmvmysql-6.0.0-alphamysqlcdmysql./configure--prefix=/usr/local--with-extra-charsets=complex--with-plugin-ndbcluster--with-plugin-partition--with-plugin-innobase--with-unix-socket-path=/usr/local/var/mysql.sockmake&&makeinstall#ln-s/usr/local/libexec/ndbd/usr/bin#ln-s/usr/local/libexec/ndb_mgmd/usr/bin#ln-s/usr/local/libexec/ndb_cpcd/usr/bin#ln-s/usr/local/libexec/mysqld/usr/bin#ln-s/usr/local/libexec/mysqlmanager/usr/bin#mysql_install_db--user=mysql我用的是:mysql-5.0.85-linux-i686.tar.gz三、安装并配置节点以下步骤需要在SerA和SerB上各做一次1.配置管理节点配置文件：mkdir/var/lib/mysql-clustercd/var/lib/mysql-clusterviconfig.ini在config.ini中添加如下内容：[ndbddefault]NoOfReplicas=2MaxNoOfConcurrentOperations=10000Amountofmemoryrequired=(SizeofDatabase*NumberOfReplicas*1.1)/NumberOfDataNodesDataMemory=128MIndexMemory=24MTimeBetweenWatchDogCheck=30000DataDir=/var/lib/mysql-clusterMaxNoOfOrderedIndexes=512StartPartialTimeout=100StartPartitionedTimeout=100ArbitrationTimeout=5000TransactionDeadlockDetectionTimeout=5000HeartbeatIntervalDbDb=5000StopOnError=0[ndb_mgmddefault]DataDir=/var/lib/mysql-cluster[ndb_mgmd]Id=1HostName=0[ndb_mgmd]Id=2HostName=[ndbd]Id=3HostName=0[ndbd]Id=4HostName=[mysqld]ArbitrationRank=2（非常重要，全靠有它,才可以形成仲裁竞争，从而当另一个机子当了时,此机还可以有知道partion完整的节点）[mysqld]ArbitrationRank=2Linux认证[mysqld]（多出的这项是留给恢复时使用的.）[mysqld]（多出的这项是留给恢复时使用的.）[tcpdefault]PortNumber=63132[separator]2.配置通用f文件，mysqld及ndbd,ndb_mgmd均使用此文件.vi/etc/f在f中添加如下内容：[mysqld]datadir=/usr/local/varsocket=/usr/local/var/mysql.sockDefaulttousingoldpasswordformatforcompatibilitywithmysql3.x#clients(thoseusingthemysqlclient10compatibilitypackage).old_passwords=1default-storage-engine=ndbclusterndbclusterndb-connectstring=0,[ndbd]connect-string=0,[ndb_mgm]connect-string=0,[ndb_mgmd]config-file=/var/lib/mysql-cluster/config.ini[mysql_cluster]ndb-connectstring=0,[mysql.server]user=mysqlbasedir=/usr/local/[mysqld_safe]log-error=/var/log/mysqld.log#pid-file=/var/run/mysqld/mysqld.pid[mysql]#socket=/usr/local/var/mysql.sock[mysqladmin]#socket=/usr/local/var/mysql.sock[ndb_restoredefault]保存退出后.四.启动管理节点SerA为：[root@SerA~]#ndb_mgmd--ndb_nodeid=1Clusterconfigurationwarning:arbitratorwithid1anddbnodewithid3onsamehost0arbitratorwithid2anddbnodewithid4onsamehostarbitratorwithid5hasnohostnamespecifiedarbitratorwithid6hasnohostnamespecifiedRunningarbitratoronthesamehostasadatabasenodemaycausecompleteclustershutdownincaseofhostfailure.注:在启动时有一个警告提示说节点1和3,2和4的arbitrator一样，可能引起整个集群失败。（可以不用放在心上）启动管理节点SerB为：[root@SerB~]#ndb_mgmd--ndb_nodeid=2五.初始化集群在SerA中[root@SerA~]#ndbd--ndb_nodeid=3--initial在SerB中[root@SerB~]#ndbd--ndb_nodeid=4--initial注：只有在第一次启动ndbd时或者对config.ini进行改动后才需要使用--initial参数！（在下面为了进行恢复实验时还再次使用到）检查工作状态在任意一台机子上启动管理终端：[root@SerA~]#ndb_mgm-eshowConnectedtoManage