Mysql高可用架構MHA搭建及測試故障轉移

好了，錯誤解決了，下面我們再次重復上面步驟:
master上mysql服務：shell> /etc/init.d/mysqld stop
再次查看manager機子上監控文件內容

shell> tail -f tail -f /masterha/app1/manager.log
日志文件顯示：
-----------------------------------------------------------
----- Failover Report -----
app1: MySQL Master failover 10.10.54.156 to 10.10.54.155 succeeded
Master 10.10.54.156 is down!
Check MHA Manager logs at mycentos4:/masterha/app1/manager.log for details.
Started automated(non-interactive) failover.
The latest slave 10.10.54.155(10.10.54.155:3306) has all relay logs for recovery.
Selected 10.10.54.155 as a new master.
10.10.54.155: OK: Applying all logs succeeded.
10.10.54.157: This host has the latest relay log events.
Generating relay diff files from the latest slave succeeded.
10.10.54.157: OK: Applying all logs succeeded. Slave started, replicating from 10.10.54.155.
10.10.54.155: Resetting slave info succeeded.
Master failover to 10.10.54.155(10.10.54.155:3306) completed successfully.
--------------------------------------------------------

2.切換成功後，檢查replication狀態
[master:156]
shell> /etc/init.d/mysqld start
[manager:154]
shell> masterha_check_repl --conf=/etc/masterha/app1.cnf
--------------------------------------------------------------
Sun Mar 2 13:22:11 2014 - [info] Slaves settings check done.
Sun Mar 2 13:22:11 2014 - [info]
10.10.54.155 (current master)
+--10.10.54.156
+--10.10.54.157
...
MySQL Replication Health is OK.
---------------------------------------------------------------
上面的"10.10.54.155 (current master)" 這句表明master成功切換到155機子上

十.上一步測試之後，新master機為155，宕掉155機子，再次測試故障轉移

1.啟動管理節點
shell> nohup masterha_manager --conf=/etc/masterha/app1.cnf > /tmp/mha_manager.log 2>&1
2.啟動日志檢測，然後當掉新master(155)，然後查看監控文件變化
shell> tail -f /masterha/app1/manager.log
3.當掉155機子(即新的master)
shell> /etc/init.d/mysqld stop
4.查看manager主機上的監控文件變化
[error][/usr/share/perl5/vendor_perl/MHA/MasterFailover.pm, ln295] Last failover was done at 2014/03/02 13:02:47. Current time is too early to do failover again. If you want to do failover, manually remove /masterha/app1/app1.failover.complete and run this script again.
錯誤解決辦法
1.日志文件提示切換master過快，需要刪除/masterha/app1/app1.failover.complete
1.刪除app1.failover.complete
shell> rm /masterha/app1/app1.failover.complete
5.重新測試：
master轉移成功，重新轉為156機子
--------------------------------------------------------
Master 10.10.54.155 is down!
Check MHA Manager logs at mycentos4:/masterha/app1/manager.log for details.
Started automated(non-interactive) failover.
The latest slave 10.10.54.156(10.10.54.156:3306) has all relay logs for recovery.
Selected 10.10.54.156 as a new master.
10.10.54.156: OK: Applying all logs succeeded.
10.10.54.157: This host has the latest relay log events.
Generating relay diff files from the latest slave succeeded.
10.10.54.157: OK: Applying all logs succeeded. Slave started, replicating from 10.10.54.156.
10.10.54.156: Resetting slave info succeeded.
Master failover to 10.10.54.156(10.10.54.156:3306) completed successfully.
-----------------------------------------------------------

//附:故障轉移後，用命令恢復原來的master

[manager:154]
1.在舊master上執行
mysql> reset master;
mysql> change master to master_host='10.10.54.155', master_port=3306, master_user='rep', master_password='rep123', master_log_file='mysql-bin.000031', master_log_pos=112;
mysql> start slave; #暫時先把舊master變為從
2.然後在manager節點上：
[manager:154]
shell> masterha_master_switch --master_state=alive --conf=/etc/masterha/app1.cnf
##master成功切換回