MySQL Galera 集群的安装过程

MySQL Galera集群版的mysql是经过codeship打了wsrep补丁的,不是普通的mysql版本.可以从github下载。

环境:redhat 7

mysql:5.5.37

galera: 25.3.5

解压，新增mysql用户，然后初始化mysql

[root@os1 ~]# ls -l mysql-5.5.37_wsrep_25.10-linux-x86_64.tar.gz

-rw-r--r--. 1 root root 62014481 Nov 23 17:17 mysql-5.5.37_wsrep_25.10-linux-x86_64.tar.gz

[root@os1 mysql]# ./scripts/mysql_install_db --user=mysql --datadir=/usr/local/mysql/data

安装 galera的plugin

[root@os1 ~]# rpm -ivh --force galera-25.3.5-1.rhel6.x86_64.rpm

Preparing...                          ################################# [100%]

Updating / installing...

1:galera-25.3.5-1.rhel6            ################################# [100%]

拷贝libgalera的plugin到mysql/lib目录下

[root@os1 local]# cp /usr/lib64/galera/libgalera_smm.so /usr/local/mysql/lib/plugin

[root@os1 ~]# chown -R mysql:mysql /usr/local/mysql

编写/etc/my.cnf文件

[mysqld]

datadir=/usr/local/mysql/data

socket=/usr/local/mysql/mysql.sock

# Disabling symbolic-links is recommended to prevent assorted security risks

symbolic-links=0

# Settings user and group are ignored when systemd is used.

# If you need to run mysqld under a different user or group,

# customize your systemd unit file for mariadb according to the

# instructions in http://fedoraproject.org/wiki/Systemd

default_storage_engine=innodb

log_bin=1

binlog_format=ROW

innodb_file_per_table=1

log-slave-updates=1

server_id=1225

innodb_autoinc_lock_mode=2

wsrep_node_name = 192.168.12.25

wsrep_provider = /usr/local/mysql/lib/plugin/libgalera_smm.so

wsrep_sst_method = mysqldump

wsrep_sst_auth=root:rootroot  --mysql密码

然后启动第一个mysql节点

[root@os1 plugin]# /usr/local/mysql/bin/mysqld_safe --wsrep-cluster-address=gcomm://

141123 19:48:06 mysqld_safe Logging to '/var/log/mysqld.log'.

141123 19:48:06 mysqld_safe Starting mysqld daemon with databases from /usr/local/mysql

141123 19:48:06 mysqld_safe WSREP: Running position recovery with --log_error='/usr/local/mysql/wsrep_recovery.u1fVym' --pid-file='/usr/local/mysql/os1.abc.com-recover.pid'

141123 19:48:06 mysqld_safe WSREP: Failed to recover position:

''

..............

[root@os1 support-files]# netstat -nltp | grep 3306

tcp        0      0 0.0.0.0:3306            0.0.0.0:*               LISTEN      10023/mysqld

[root@os1 support-files]# /usr/local/mysql/bin/mysql -u root -p --socket=/usr/local/mysql/mysql/mysql.sock

mysql> show global status like '%state%';

+---------------------------+--------------------------------------+

| Variable_name             | Value                                |

+---------------------------+--------------------------------------+

| wsrep_local_state_uuid    | b34c90ed-7306-11e4-ba41-a65afe12d31b |

| wsrep_local_state         | 4                                    |

| wsrep_local_state_comment | Synced                               |

| wsrep_cluster_state_uuid  | b34c90ed-7306-11e4-ba41-a65afe12d31b |

+---------------------------+--------------------------------------+

在第二个节点如法炮制，但是/etc/my.cnf稍有不同

[mysqld]

datadir=/usr/local/mysql/data

socket=/usr/local/mysql/mysql.sock

# Disabling symbolic-links is recommended to prevent assorted security risks

symbolic-links=0

# Settings user and group are ignored when systemd is used.

# If you need to run mysqld under a different user or group,

# customize your systemd unit file for mariadb according to the

# instructions in http://fedoraproject.org/wiki/Systemd

default_storage_engine=innodb

log_bin=1

binlog_format=ROW

innodb_file_per_table=1

log-slave-updates=1

server_id=12213

innodb_autoinc_lock_mode=2

wsrep_node_name = 192.168.12.213

wsrep_provider = /usr/local/mysql/lib/plugin/libgalera_smm.so

wsrep_sst_method = mysqldump

wsrep_sst_auth=root:rootroot

在第二个节点上启动mysql

[root@os2 mysql]# /usr/local/mysql/bin/mysqld_safe --wsrep-cluster-address="gcomm://192.168.12.25:4567"

.........

141123 20:08:42 [Note] WSREP: Setting initial position to 00000000-0000-0000-0000-000000000000:-1

141123 20:08:42 [Note] WSREP: protonet asio version 0

141123 20:08:42 [Note] WSREP: Using CRC-32C (optimized) for message checksums.

141123 20:08:42 [Note] WSREP: backend: asio

141123 20:08:42 [Note] WSREP: GMCast version 0

141123 20:08:42 [Note] WSREP: (77c88cef-7309-11e4-8435-43850781c6da, 'tcp://0.0.0.0:4567') listening at tcp://0.0.0.0:4567

141123 20:08:42 [Note] WSREP: (77c88cef-7309-11e4-8435-43850781c6da, 'tcp://0.0.0.0:4567') multicast: , ttl: 1

141123 20:08:42 [Note] WSREP: EVS version 0

141123 20:08:42 [Note] WSREP: PC version 0

141123 20:08:42 [Note] WSREP: gcomm: connecting to group 'my_wsrep_cluster', peer '192.168.12.25:4567'

141123 20:08:45 [Warning] WSREP: no nodes coming from prim view, prim not possible

从第一个节点上查看，wsrep_cluser相关变量，表明集群现在已经就绪.

mysql> show global status like '%wsrep_cluster%';

+--------------------------+--------------------------------------+

| Variable_name            | Value                                |

+--------------------------+--------------------------------------+

| wsrep_cluster_conf_id    | 2                                    |

| wsrep_cluster_size       | 2                                    |

| wsrep_cluster_state_uuid | ff4f28f7-7318-11e4-8e26-8bcedbe58a6d |

| wsrep_cluster_status     | Primary                              |

+--------------------------+--------------------------------------+

在两个节点上相互增加数据测试，可以见到对方数据，表明集群正常. 如果是第一次搭建，也许会碰到很多坑，

作者打算写专门一个文章来总结.