1.概述
本篇讲解group replication的安装部署。由于group replication最大支持9个节点，本文将大家在一台机器上快速部署9节点的group replication，多机器部署是一样的,没有任何区别。
2.安装规划
group_replication_group_name=”aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa”
节点名 Internal port Db port datadir conf mode s1 5701 9001 /data/for_gr/s1 s1.cnf read/write s2 5702 9002 /data/for_gr/s2 S2.cnf read/write s3 5703 9003 /data/for_gr/s3 S3.cnf read/write s4 5704 9004 /data/for_gr/s4 S4.cnf read/write s5 5705 9005 /data/for_gr/s5 S5.cnf read/write s6 5706 9006 /data/for_gr/s6 S6.cnf read/write s7 5707 9007 /data/for_gr/s7 S7.cnf read/write s8 5708 9008 /data/for_gr/s8 S8.cnf read/write s9 5709 9009 /data/for_gr/s9 S9.cnf read/write
3.目录和二进制文件准备
3.1二进制安装包准备
直接在官网下载5.7.17，解压放到/opt下
[mysql@10-25-126-227 mysql-5.7.17]$ pwd /opt/mysql-5.7.17 [mysql@10-25-126-227 mysql-5.7.17]$ ls -l total 52 drwxrwxr-x 2 mysql mysql 4096 Dec 13 11:44 bin -rw-r--r-- 1 mysql mysql 17987 Nov 28 21:32 COPYING drwxrwxr-x 2 mysql mysql 4096 Dec 13 11:44 docs drwxrwxr-x 3 mysql mysql 4096 Dec 13 11:44 include drwxrwxr-x 5 mysql mysql 4096 Dec 13 11:44 lib drwxrwxr-x 4 mysql mysql 4096 Dec 13 11:44 man -rw-r--r-- 1 mysql mysql 2478 Nov 28 21:32 README drwxrwxr-x 28 mysql mysql 4096 Dec 13 11:44 share drwxrwxr-x 2 mysql mysql 4096 Dec 13 11:44 support-files
3.2数据目录准备
mkdir -p /data/for_gr/{s1,s2,s3,s4,s5,s6,s7,s8,s9}
3.3配置文件准备
这里以s1配置为例，其他配置文件修改数据目录和端口即可。
[mysqld] # server configuration datadir=/data/for_gr/s1 #不同节点，数据目录修改即可 basedir=/opt/mysql-5.7.17 port=9001 #修改为响应节点的端口，用户客户端连接 socket=/data/for_gr/s1/s1.sock #socket也改到响应数据目录 server_id=1 #server-id别忘记改，就1-9吧 gtid_mode=ON enforce_gtid_consistency=ON master_info_repository=TABLE relay_log_info_repository=TABLE binlog_checksum=NONE log_slave_updates=ON log_bin=binlog binlog_format=ROW innodb_buffer_pool_instances=4 innodb_buffer_pool_size=1G innodb_flush_log_at_trx_commit=2 sync_binlog=0 #for parallel apply binlog slave-parallel-type=LOGICAL_CLOCK slave-parallel-workers=4 slave_preserve_commit_order=on #for group replication transaction_write_set_extraction=XXHASH64 loose-group_replication_group_name="aaaaaaaa-aaaa-aaaa-aaaa-aaaaaaaaaaaa" loose-group_replication_start_on_boot=off loose-group_replication_local_address= "127.0.0.1:5701" #改一下端口就可以了 loose-group_replication_group_seeds= "127.0.0.1:5701,127.0.0.1:5702,127.0.0.1:5703,127.0.0.1:5704,127.0.0.1:5705,127.0.0.1:5706,127.0.0.1:5707,127.0.0.1:5708,127.0.0.1:5709" loose-group_replication_bootstrap_group= off loose-group_replication_single_primary_mode=false
其他节点，改一下上面注释部分就可以了。
复制配置文件:
for i in {2..9};do cp s1.cnf s$i.cnf;done
修改相关配置:
for i in {2..9};do sed -i "s/s1/s$i/g" s$i.cnf;done
for i in {2..9};do sed -i "s/server_id=1/server_id=$i/g" s$i.cnf;done
for i in {2..9};do sed -i "s/port=9001/port=900$i/g" s$i.cnf;done
for i in {2..9};do sed -i "s/loose-group_replication_local_address= \"127.0.0.1:5701\"/loose-group_replication_local_address= \"127.0.0.1:570$i\"/g" s$i.cnf;done
确认准备好的目录和配置文件:
[mysql@10-25-126-227 for_gr]$ tree /data/conf/for_gr/ /data/conf/for_gr/ ├── s1.cnf ├── s2.cnf ├── s3.cnf ├── s4.cnf ├── s5.cnf ├── s6.cnf ├── s7.cnf ├── s8.cnf └── s9.cnf 0 directories, 9 files [mysql@10-25-126-227 for_gr]$ tree /data/for_gr/ /data/for_gr/ ├── s1 ├── s2 ├── s3 ├── s4 ├── s5 ├── s6 ├── s7 ├── s8 └── s9 9 directories, 0 files
4.初始化数据库
for i in {1..9};do /opt/mysql-5.7.17/bin/mysqld --initialize-insecure --basedir=/opt/mysql-5.7.17 --datadir=/data/for_gr/s$i ;done;
5.启动数据库
[mysql@10-25-126-227 mysql-5.7.17]$ cd /opt/mysql-5.7.17/
[mysql@10-25-126-227 mysql-5.7.17]$ pwd
/opt/mysql-5.7.17
[mysql@10-25-126-227 mysql-5.7.17]$ for i in {1..9}; do ./bin/mysqld_safe --defaults-file=/data/conf/for_gr/s$i.cnf & done
[mysql@10-25-126-227 mysql-5.7.17]$ pidof mysqld
5615 5608 5607 5606 5604 5600 5592 5591 5585
6.添加复制用户
由于group replication只支持innodb引擎，用户放在mysql.user里面，是myisam引擎，所以我们先把每个节点的用户准备好，分节点创建。这里我们创建两个用户,一个用于group replication内部使用，另外一个用户管理整个group replication。
注意:添加用户不要写到binlog去。
6.1准备好用户创建脚本
[mysql@10-25-126-227 ~]$ cat ~/user.sql SET SQL_LOG_BIN=0; CREATE USER rpl_user@'%'; GRANT REPLICATION SLAVE ON *.* TO rpl_user@'%' IDENTIFIED BY 'rpl_pass'; CREATE USER dbmon@'%'; GRANT ALL PRIVILEGES ON *.* TO dbmon@'%' IDENTIFIED BY 'dbmon'; FLUSH PRIVILEGES; SET SQL_LOG_BIN=1; CHANGE MASTER TO MASTER_USER='rpl_user', MASTER_PASSWORD='rpl_pass' FOR CHANNEL 'group_replication_recovery';
6.2所有节点都创建一遍
cd /opt/mysql-5.7.17
for i in {1..9};do ./bin/mysql -S /data/for_gr/s$i/s$i.sock -e "source /home/mysql/user.sql";done
7.初始化group replication
注意创建一个group replication，需要在一个节点初始化，也只需要在一个节点初始化，不可在多个节点都执行。这里我们选择第一个。
[mysql@10-25-126-227 mysql-5.7.17]$ ./bin/mysql -S /data/for_gr/s1/s1.sock Welcome to the MySQL monitor. Commands end with ; or \g. Your MySQL connection id is 16 Server version: 5.7.17-log MySQL Community Server (GPL) Copyright (c) 2000, 2016, Oracle and/or its affiliates. All rights reserved. Oracle is a registered trademark of Oracle Corporation and/or its affiliates. Other names may be trademarks of their respective owners. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement. root@localhost [(none)] 03:22:30>>>INSTALL PLUGIN group_replication SONAME 'group_replication.so'; Query OK, 0 rows affected (0.00 sec) root@localhost [(none)] 03:22:54>>>SET GLOBAL group_replication_bootstrap_group=ON; Query OK, 0 rows affected (0.00 sec) root@localhost [(none)] 03:22:54>>>START GROUP_REPLICATION; Query OK, 0 rows affected (1.05 sec) root@localhost [(none)] 03:22:55>>>SET GLOBAL group_replication_bootstrap_group=OFF; Query OK, 0 rows affected (0.00 sec) root@localhost [(none)] 03:23:04>>>SELECT * FROM performance_schema.replication_group_members \G *************************** 1. row *************************** CHANNEL_NAME: group_replication_applier MEMBER_ID: e0c9963b-c4ee-11e6-99a1-525400e71e7c MEMBER_HOST: 10-25-126-227 MEMBER_PORT: 9001 MEMBER_STATE: ONLINE 1 row in set (0.00 sec)
由以上查询可以看到，我们整个group replication已经有一个节点在线了。
8.添加其他节点加入group replication
for i in {2..9};do ./bin/mysql -S /data/for_gr/s$i/s$i.sock -e "INSTALL PLUGIN group_replication SONAME 'group_replication.so';START GROUP_REPLICATION;";done
注意:这次只添加了2-9，1号节点已经在线了。
9.在线节点状态查询
九个节点都是Online状态。
10.关闭group replication
for i in {1..9};do ./bin/mysql -S /data/for_gr/s$i/s$i.sock -e "STOP GROUP_REPLICATION;";done
然后关闭所有数据库就行了。
11.启动group replication
首先按照正常流程启动数据库。
11.1启动第一个节点
启动节点首先需要在某个节点执行初始化动作，这个节点最好是最后关闭的那个节点，这里我们用S9:
root@localhost [(none)] 03:30:25>>>show variables like '%server_id'; +---------------+-------+ | Variable_name | Value | +---------------+-------+ | server_id | 9 | +---------------+-------+ 1 row in set (0.00 sec) root@localhost [(none)] 03:30:29>>>SET GLOBAL group_replication_bootstrap_group=ON; Query OK, 0 rows affected (0.00 sec) root@localhost [(none)] 03:30:38>>>START GROUP_REPLICATION; Query OK, 0 rows affected (1.01 sec) root@localhost [(none)] 03:30:39>>>SET GLOBAL group_replication_bootstrap_group=OFF; Query OK, 0 rows affected (0.00 sec) root@localhost [(none)] 03:30:55>>>SELECT MEMBER_HOST,MEMBER_PORT,MEMBER_STATE FROM performance_schema.replication_group_members; +---------------+-------------+--------------+ | MEMBER_HOST | MEMBER_PORT | MEMBER_STATE | +---------------+-------------+--------------+ | 10-25-126-227 | 9009 | ONLINE | +---------------+-------------+--------------+ 1 row in set (0.00 sec)
11.2启动添加剩下8个节点
for i in {1..8};do ./bin/mysql -S /data/for_gr/s$i/s$i.sock -e "START GROUP_REPLICATION;";done
请问在不同物理机上部署时，s2加入时，有遇到[GCS] Timeout while waiting for the group communication engine to be ready!吗？
错误日志：
2017-04-12T14:37:48.052080Z 0 [ERROR] Plugin group_replication reported: ‘[GCS] Timeout while waiting for the group communication engine to be ready!’
2017-04-12T14:37:48.052200Z 0 [ERROR] Plugin group_replication reported: ‘[GCS] The group communication engine is not ready for the member to join. Local port: 24902’
是不是你的防火墙啥的，网络通不？