利用keepalived構建高可用MySQL-HA
關于MySQL-HA,目前有多種解決方案,比如heartbeat、drbd、mmm、共享存儲,但是它們各有優缺點。heartbeat、drbd配置較為復雜,需要自己寫腳本才能實現MySQL自動切換,對于不會腳本語言的人來說,這無疑是一種腦裂問題;對于mmm,生產環境中很少有人用,且mmm 管理端需要單獨運行一臺服務器上,要是想實現高可用,就得對mmm管理端做HA,這樣無疑又增加了硬件開支;對于共享存儲,個人覺得MySQL數據還是放在本地較為安全,存儲設備畢竟存在單點隱患。
使用MySQL雙master+keepalived是一種非常好的解決方案,在MySQL-HA環境中,MySQL互為主從關系,這樣就保證了兩臺MySQL數據的一致性,然后用keepalived實現虛擬IP,通過keepalived自帶的服務監控功能來實現MySQL故障時自動切換。
下面,我把即將上線的一個生產環境中的架構與大家分享一下,看一下這個架構中,MySQL-HA是如何實現的,環境拓撲如下
2. MySQL-master1:192.168.1.201
3. MySQL-master2:192.168.1.202
4.
5. OS版本:CentOS 5.4
6. MySQL版本:5.0.89
7. Keepalived版本:1.1.20
一、MySQL master-master配置
1、修改MySQL配置文件
兩臺MySQL均如要開啟binlog日志功能,開啟方法:在MySQL配置文件[MySQLd]段中加上log-bin=MySQL-bin選項
兩臺MySQL的server-ID不能一樣,默認情況下兩臺MySQL的serverID都是1,需將其中一臺修改為2即可
2、將192.168.1.201設為192.168.1.202的主服務器
在192.168.1.201上新建授權用戶
2. Query OK, 0 rows affected (0.00 sec)
3.
4. MySQL> show master status;
5. +------------------+----------+--------------+------------------+
6. | File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
7. +------------------+----------+--------------+------------------+
8. | MySQL-bin.000003 | 374 | | |
9. +------------------+----------+--------------+------------------+
10. 1 row in set (0.00 sec)
在192.168.1.202上將192.168.1.201設為自己的主服務器
2. Query OK, 0 rows affected (0.05 sec)
3.
4. MySQL> start slave;
5. Query OK, 0 rows affected (0.00 sec)
6.
7. MySQL> show slave status\G
8. *************************** 1. row ***************************
9. Slave_IO_State: Waiting for master to send event
10. Master_Host: 192.168.1.201
11. Master_User: replication
12. Master_Port: 3306
13. Connect_Retry: 60
14. Master_Log_File: MySQL-bin.000003
15. Read_Master_Log_Pos: 374
16. Relay_Log_File: MySQL-master2-relay-bin.000002
17. Relay_Log_Pos: 235
18. Relay_Master_Log_File: MySQL-bin.000003
19. Slave_IO_Running: Yes
20. Slave_SQL_Running: Yes
21. Replicate_Do_DB:
22. Replicate_Ignore_DB:
23. Replicate_Do_Table:
24. Replicate_Ignore_Table:
25. Replicate_Wild_Do_Table:
26. Replicate_Wild_Ignore_Table:
27. Last_Errno: 0
28. Last_Error:
29. Skip_Counter: 0
30. Exec_Master_Log_Pos: 374
31. Relay_Log_Space: 235
32. Until_Condition: None
33. Until_Log_File:
34. Until_Log_Pos: 0
35. Master_SSL_Allowed: No
36. Master_SSL_CA_File:
37. Master_SSL_CA_Path:
38. Master_SSL_Cert:
39. Master_SSL_Cipher:
40. Master_SSL_Key:
41. Seconds_Behind_Master: 0
42. 1 row in set (0.00 sec)
3、將192.168.1.202設為192.168.1.201的主服務器
在192.168.1.202上新建授權用戶
2. Query OK, 0 rows affected (0.00 sec)
3.
4. MySQL> show master status;
5. +------------------+----------+--------------+------------------+
6. | File | Position | Binlog_Do_DB | Binlog_Ignore_DB |
7. +------------------+----------+--------------+------------------+
8. | MySQL-bin.000003 | 374 | | |
9. +------------------+----------+--------------+------------------+
10. 1 row in set (0.00 sec)
在192.168.1.201上,將192.168.1.202設為自己的主服務器
2. Query OK, 0 rows affected (0.05 sec)
3.
4. MySQL> start slave;
5. Query OK, 0 rows affected (0.00 sec)
6.
7. MySQL> show slave status\G
8. *************************** 1. row ***************************
9. Slave_IO_State: Waiting for master to send event
10. Master_Host: 192.168.1.202
11. Master_User: replication
12. Master_Port: 3306
13. Connect_Retry: 60
14. Master_Log_File: MySQL-bin.000003
15. Read_Master_Log_Pos: 374
16. Relay_Log_File: MySQL-master1-relay-bin.000002
17. Relay_Log_Pos: 235
18. Relay_Master_Log_File: MySQL-bin.000003
19. Slave_IO_Running: Yes
20. Slave_SQL_Running: Yes
21. Replicate_Do_DB:
22. Replicate_Ignore_DB:
23. Replicate_Do_Table:
24. Replicate_Ignore_Table:
25. Replicate_Wild_Do_Table:
26. Replicate_Wild_Ignore_Table:
27. Last_Errno: 0
28. Last_Error:
29. Skip_Counter: 0
30. Exec_Master_Log_Pos: 374
31. Relay_Log_Space: 235
32. Until_Condition: None
33. Until_Log_File:
34. Until_Log_Pos: 0
35. Master_SSL_Allowed: No
36. Master_SSL_CA_File:
37. Master_SSL_CA_Path:
38. Master_SSL_Cert:
39. Master_SSL_Cipher:
40. Master_SSL_Key:
41. Seconds_Behind_Master: 0
42. 1 row in set (0.00 sec)
4、MySQL同步測試
如上述均正確配置,現在任何一臺MySQL上更新數據都會同步到另一臺MySQL,MySQL同步在此不再演示
二、keepalived安裝及配置
1、192.168.1.201服務器上keepalived安裝及配置
安裝keepalived
2. #cd keepalived-1.1.20
3. #./configure --prefix=/usr/local/keepalived --with-kernel-dir=/usr/src/kernels/2.6.18-164.el5-i686
4. #make && make install
配置keepalived
我們自己在新建一個配置文件,默認情況下keepalived啟動時會去/etc/keepalived目錄下找配置文件
2. #vi /etc/keepalived/keepalived.conf
3. ! Configuration File for keepalived
4. global_defs {
5. notification_email {
6. luwenju@live.cn
7. }
8. notification_email_from luwenju@live.cn
9. smtp_server 127.0.0.1
10. smtp_connect_timeout 30
11. router_id MySQL-ha
12. }
13.
14. vrrp_instance VI_1 {
15. state BACKUP #兩臺配置此處均是BACKUP
16. interface eth0
17. virtual_router_id 51
18. priority 100 #優先級,另一臺改為90
19. advert_int 1
20. nopreempt #不搶占,只在優先級高的機器上設置即可,優先級低的機器不設置
21. authentication {
22. auth_type PASS
23. auth_pass 1111
24. }
25. virtual_ipaddress {
26. 192.168.1.200
27. }
28. }
29.
30. virtual_server 192.168.1.200 3306 {
31. delay_loop 2 #每個2秒檢查一次real_server狀態
32. lb_algo wrr #LVS算法
33. lb_kind DR #LVS模式
34. persistence_timeout 60 #會話保持時間
35. protocol TCP
36. real_server 192.168.1.201 3306 {
37. weight 3
38. notify_down /usr/local/MySQL/bin/MySQL.sh #檢測到服務down后執行的腳本
39. TCP_CHECK {
40. connect_timeout 10 #連接超時時間
41. nb_get_retry 3 #重連次數
42. delay_before_retry 3 #重連間隔時間
43. connect_port 3306 #健康檢查端口
44. }
45. }
編寫檢測服務down后所要執行的腳本
#!/bin/sh
pkill keepalived
#chmod +x /usr/local/MySQL/bin/MySQL.sh
注:此腳本是上面配置文件notify_down選項所用到的,keepalived使用notify_down選項來檢查real_server的服務狀態,當發現real_server服務故障時,便觸發此腳本;我們可以看到,腳本就一個命令,通過pkill keepalived強制殺死keepalived進程,從而實現了MySQL故障自動轉移。另外,我們不用擔心兩個MySQL會同時提供數據更新操作,因為每臺MySQL上的keepalived的配置里面只有本機MySQL的IP+VIP,而不是兩臺MySQL的IP+VIP
啟動keepalived
2. #ps -aux | grep keepalived
測試
找一臺局域網PC,然后去ping MySQL的VIP,這時候MySQL的VIP是可以ping的通的
停止MySQL服務,看keepalived健康檢查程序是否會觸發我們編寫的腳本
2、192.168.1.202上keepalived安裝及配置
安裝keepalived
2. #cd keepalived-1.1.20
3. #./configure --prefix=/usr/local/keepalived --with-kernel-dir=/usr/src/kernels/2.6.18-164.el5-i686
4. #make && make install
配置keepalived
這臺配置和上面基本一樣,但有三個地方不同:優先級為90、無搶占設置、real_server為本機IP
2. #vi /etc/keepalived/keepalived.conf
3. ! Configuration File for keepalived
4. global_defs {
5. notification_email {
6. luwenju@live.cn
7. }
8. notification_email_from luwenju@live.cn
9. smtp_server 127.0.0.1
10. smtp_connect_timeout 30
11. router_id MySQL-ha
12. }
13.
14. vrrp_instance VI_1 {
15. state BACKUP
16. interface eth0
17. virtual_router_id 51
18. priority 90
19. advert_int 1
20. authentication {
21. auth_type PASS
22. auth_pass 1111
23. }
24. virtual_ipaddress {
25. 192.168.1.200
26. }
27. }
28.
29. virtual_server 192.168.1.200 3306 {
30. delay_loop 2
31. lb_algo wrr
32. lb_kind DR
33. persistence_timeout 60
34. protocol TCP
35. real_server 192.168.1.202 3306 {
36. weight 3
37. notify_down /usr/local/MySQL/bin/MySQL.sh
38. TCP_CHECK {
39. connect_timeout 10
40. nb_get_retry 3
41. delay_before_retry 3
42. connect_port 3306
43. }
44. }
編寫檢測服務down后所要執行的腳本
2. #!/bin/sh
3. pkill keepalived
4. #chmod +x /usr/local/MySQL/bin/MySQL.sh
5.
6. 啟動keepalived
7. #/usr/local/keepalived/sbin/keepalived –D
8. #ps -aux | grep keepalived
測試
停止MySQL服務,看keepalived健康檢查程序是否會觸發我們編寫的腳本
三、測試
MySQL遠程登錄測試
我們找一臺安裝有MySQL客戶端的windows,然后登錄VIP,看是否能登錄,在登錄之兩臺MySQL服務器都要授權允許從遠程登錄
2. Query OK, 0 rows affected (0.00 sec)
3.
4. MySQL> flush privileges;
5. Query OK, 0 rows affected (0.00 sec)
使用客戶端登錄VIP測試
2. Welcome to the MySQL monitor. Commands end with ; or \g.
3. Your MySQL connection id is 224
4. Server version: 5.0.89-log Source distribution
5.
6. Type 'help;' or '\h' for help. Type '\c' to clear the current input statement.
7.
8. MySQL>
● keepalived故障轉移測試
※在windows客戶端一直去ping VIP,然后關閉192.168.1.201上的keepalived,正常情況下VIP就會切換到192.168.1.202上面去
※開啟192.168.1.201上的keepalived,關閉192.168.1.202上的keepalived,看是否能自動切換,正常情況下VIP又會屬于192.168.1.201
注:keepalived切換速度還是非常塊的,整個切換過程只需1-3秒
● MySQL故障轉移測試
※在192.168.1.201上關閉MySQL服務,看VIP是否會切換到192.168.1.202上
※開啟192.168.1.201上的MySQL和keepalived,然后關閉192.168.1.202上的MySQL,看VIP是否會切換到192.168.1.201上
下面是用windows客戶端連接的MySQL的VIP,在切換時我執行了一個MySQL查詢命令,從執行show databases到顯示出結果時間為3-5秒(大家可以看到上面有個錯誤提示,不過不用擔心,因為我們的keepalived切換大概為3秒左右,這3秒左右VIP是誰都不屬于的)
2. ERROR 2006 (HY000): MySQL server has gone away
3. No connection. Trying to reconnect...
4. Connection id: 592
5. Current database: *** NONE ***
6.
7. +--------------------+
8. | Database |
9. +--------------------+
10. | information_schema |
11. | MySQL |
12. | test |
13. +--------------------+
14. 3 rows in set (9.01 sec)
后話:世間萬事萬物,都不具備絕對的完美,就像上面的MySQL-HA一樣,keepalived只能做到對3306的健康檢查,但是做不到比如像MySQL復制中的slave-SQL、slave-IO進程的檢查。所以要想做到一些細致的健康檢查,還得需要借助額外的監控工具,比如nagios,然后用nagios實現短信、郵件報警,從而能夠有效地解決問題。