环境概述
此文将搭建一个两节点的fabric HA集群
角色
IP
port
备注
Fabric
localhost
3306
管理节点
node1
localhost
3307
HA成员
node1
localhost
3308
HA成员
安装Fabric Fabric现在已经合并到 utilities 中了，但是，下载注意官方的提示：MySQL Fabric is included in MySQL Utilities versions prior to 1.6.2. https://cdn.mysql.com//Downloads/MySQLGUITools/mysql-utilities-1.5.6-1.el6.noarch.rpm
MySQL配置 重点设置下面四个参数
1 2 3 4 log_bin gtid-mode=ON enforce-gtid-consistency log_slave_updates
Fabric配置 创建账号 每个节点都要创建
1 2 3 create user 'fabric' @'%' identified by '123' ; grant all on *.* to 'fabric' @'%' ; flush privileges;
修改MySQL Fabric 配置文件 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 vim /etc/mysql/fabric.cfg [storage] address = 192.168.1.100:3306 user = fabric password = pass database = fabric auth_plugin = mysql_native_password connection_timeout = 6 connection_attempts = 6 connection_delay = 1 [servers] user = fabric password = pass unreachable_timeout = 5
HA初始化 1 2 3 4 5 6 7 8 9 [root@localhost ~] [INFO] 1493595918.431773 - MainThread - Initializing persister: user (fabric), server (192.168.1.201:3306), database (fabric). Finishing initial setup ======================= Password for admin user is not yet set . Password for admin/xmlrpc: Repeat Password: Password set . Password set .
创建HA集群 创建group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 [root@localhost ~] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 1f8d4b07-6d84-48be-9178-da695397af3c 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x10ad310>. 4 2 1.4936e+09 Executing action (_create_group). 5 2 1.4936e+09 Executed action (_create_group).
添加成员到group 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 [root@localhost ~] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 87ea2237-9c81-4cdb-82d4-e88c2028046f 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x10ad6d0>. 4 2 1.4936e+09 Executing action (_add_server). 5 2 1.4936e+09 Executed action (_add_server). [root@localhost ~] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 91bff57f-bc1d-4749-89a1-95183f932900 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x10ad6d0>. 4 2 1.4936e+09 Executing action (_add_server). 5 2 1.4936e+09 Executed action (_add_server).
提升一个主 有两种方式提升一个成员为主，一是让fabric自己选择，而是手动指定。
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 [root@localhost ~] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ ae00bc31-e27f-42d9-a656-bad4c0cd2921 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0xec4690>. 4 2 1.4936e+09 Executing action (_define_ha_operation). 5 2 1.4936e+09 Executed action (_define_ha_operation). 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x101ae90>. 4 2 1.4936e+09 Executing action (_find_candidate_fail). 5 2 1.4936e+09 Executed action (_find_candidate_fail). 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x101ac10>. 4 2 1.4936e+09 Executing action (_check_candidate_fail). 5 2 1.4936e+09 Executed action (_check_candidate_fail). 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x101ab50>. 4 2 1.4936e+09 Executing action (_wait_slave_fail). 5 2 1.4936e+09 Executed action (_wait_slave_fail). 3 2 1.4936e+09 Triggered by <mysql.fabric.events.Event object at 0x1028110>. 4 2 1.4936e+09 Executing action (_change_to_candidate). 5 2 1.4936e+09 Executed action (_change_to_candidate).
1 mysqlfabric group promote <group_name> --slave_id='<node_uuid>'
HA健康检查 两种方式，一种是UUID显示方式：
1 2 3 4 5 6 7 8 9 [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid is_alive status is_not_running is_not_configured io_not_running sql_not_running io_error sql_error ------------------------------------ -------- --------- -------------- ----------------- -------------- --------------- -------- --------- f0c7d668-2a7c-11e7-b3c3-000c29217b03 1 PRIMARY 0 0 0 0 False False fe6d0231-2a7c-11e7-b65e-000c29217b03 1 SECONDARY 0 0 0 0 False False
另一种是IP显示方式:
1 2 3 4 5 6 7 8 9 [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 server_uuid address status mode weight ------------------------------------ ------------------ --------- ---------- ------ f0c7d668-2a7c-11e7-b3c3-000c29217b03 192.168.1.201:3307 PRIMARY READ_WRITE 1.0 fe6d0231-2a7c-11e7-b65e-000c29217b03 192.168.1.201:3308 SECONDARY READ_ONLY 1.0
自动故障转移 激活故障探测 整个集群已经搭建完毕，接下来我们要让集群具有发现问题的能力，
1 2 3 4 5 6 7 8 9 10 11 12 13 14 [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 39d80744-c811-4e24-a2c8-2b8a5a9005fa 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.49365e+09 Triggered by <mysql.fabric.events.Event object at 0x101aa90>. 4 2 1.49365e+09 Executing action (_activate_group). 5 2 1.49365e+09 Executed action (_activate_group).
模拟故障 现在 master=3307 slave=3308,我们手动关闭3307，看HA故障转移能力
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 [root@localhost data3307] [root@localhost data3307] Reporting MySQL servers MySQL server from group: mysqld3306 is running MySQL server from group: mysqld3307 is not running MySQL server from group: mysqld3308 is running MySQL server from group: mysqld3309 is running [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 server_uuid address status mode weight ------------------------------------ ------------------ ------- ---------- ------ f0c7d668-2a7c-11e7-b3c3-000c29217b03 192.168.1.201:3307 FAULTY READ_WRITE 1.0 fe6d0231-2a7c-11e7-b65e-000c29217b03 192.168.1.201:3308 PRIMARY READ_WRITE 1.0
可以看到master已经转移到了3308 。But，当我们重启3307之后，3307并不能自动加入集群，需要我们手动做些工作。
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 3a2ac6b2-5a79-4ab8-9d3d-a197332ee711 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.49365e+09 Triggered by <mysql.fabric.events.Event object at 0x10ada50>. 4 2 1.49365e+09 Executing action (_set_server_status). 5 2 1.49365e+09 Executed action (_set_server_status). [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 uuid finished success result ------------------------------------ -------- ------- ------ 8fa3a763-aa48-489d-95a4-52f5323d0bbc 1 1 1 state success when description ----- ------- ------------- ------------------------------------------------------------- 3 2 1.49365e+09 Triggered by <mysql.fabric.events.Event object at 0x10ada50>. 4 2 1.49365e+09 Executing action (_set_server_status). 5 2 1.49365e+09 Executed action (_set_server_status).
接着查看集群状态：
1 2 3 4 5 6 7 8 9 [root@localhost data3307] Password for admin: Fabric UUID: 5ca1ab1e-a007-feed-f00d-cab3fe13249e Time-To-Live: 1 server_uuid address status mode weight ------------------------------------ ------------------ --------- ---------- ------ f0c7d668-2a7c-11e7-b3c3-000c29217b03 192.168.1.201:3307 SECONDARY READ_ONLY 1.0 fe6d0231-2a7c-11e7-b65e-000c29217b03 192.168.1.201:3308 PRIMARY READ_WRITE 1.0
至此，3307已经重新加入ha，成为slave。以此类推，将其他节点下线，再上线，也能达到同样效果，不再演示。至此，我们模拟了两个节点分别停机，上线，fabric的自动故障转移功能。
配置过程的报错 问题1 Q:在提升主的时候，由于要选择的主此前purge掉了一些binlog 报错：
1 2 3 Got fatal error 1236 from master when reading data from binary log : 'The slave is connecting \ using CHANGE MASTER TO MASTER_AUTO_POSITION = 1, but the master has purged binary logs\ containing GTIDs that the slave requires.'
A:手动将slave指向master
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 master: mysql> show master status\G; *************************** 1. row *************************** File: mysql-bin.000025 Position: 393 Binlog_Do_DB: Binlog_Ignore_DB: Executed_Gtid_Set: d54fe35a-2a7c-11e7-b24b-000c29217b03:1-73, f0c7d668-2a7c-11e7-b3c3-000c29217b03:1-16 1 row in set (0.00 sec) slave: mysql> stop slave ; Query OK, 0 rows affected (0.01 sec) mysql> reset slave ; Query OK, 0 rows affected (0.09 sec) mysql> reset master; Query OK, 0 rows affected (0.02 sec) mysql> set global gtid_purged="d54fe35a-2a7c-11e7-b24b-000c29217b03:1-73,f0c7d668-2a7c-11e7-b3c3-000c29217b03:1-16" ; Query OK, 0 rows affected (0.00 sec) mysql> change master to master_host='localhost' ,master_port=3307,master_user='fabric' ,master_password='123' ,master_auto_position=1; Query OK, 0 rows affected, 2 warnings (0.05 sec) mysql> start slave; Query OK, 0 rows affected (0.02 sec)
问题2 Q:在上个问题的处理过程中，曾试图跳过部分事务，所以设置了GTID_NEXT参数，然后就出现如下报错
1 2 3 4 5 6 7 mysql> stop slave ; ERROR 1837 (HY000): When @@SESSION.GTID_NEXT is set to a GTID, you must explicitly set it to a different value after a COMMIT or ROLLBACK. Please check GTID_NEXT variable manual page for detailed explanation. Current @@SESSION.GTID_NEXT is 'f0c7d668-2a7c-11e7-b3c3-000c29217b03:11' . mysql> start slave ; ERROR 1837 (HY000): When @@SESSION.GTID_NEXT is set to a GTID, you must explicitly set it to a different value after a COMMIT or ROLLBACK. Please check GTID_NEXT variable manual page for detailed explanation. Current @@SESSION.GTID_NEXT is 'f0c7d668-2a7c-11e7-b3c3-000c29217b03:11' . mysql> reset slave ; ERROR 1837 (HY000): When @@SESSION.GTID_NEXT is set to a GTID, you must explicitly set it to a different value after a COMMIT or ROLLBACK. Please check GTID_NEXT variable manual page for detailed explanation. Current @@SESSION.GTID_NEXT is 'f0c7d668-2a7c-11e7-b3c3-000c29217b03:11' .
A:
1 2 mysql> set gtid_next='automatic' ; Query OK, 0 rows affected (0.01 sec)
问题3 Q:当3307下线，从新上线后，设置spare的时候，报错
1 Last_SQL_Error: Slave failed to initialize relay log info structure from the repository
** A:**
