Atlas实现读写分离
环境介绍：
试验机器4台：
192.168.0.41 Atlas
192.168.0.42 Master
192.168.0.43 Slave-1
192.168.0.44 Slave-2
4台机器均为RHEL 6.4 x86_64系统
[[email protected] ~]# uname -a
Linux Nginx 2.6.32-358.el6.x86_64 #1 SMP Tue Jan 29 11:47:41 EST 2013 x86_64 x86_64 x86_64 GNU/Linux
MySQL版本为5.5.21
防火墙及SElinux全部关闭；无VIP；
试验机的主从已搭建完毕，数据一致；
安装情况：
安装包使用的是RPM包
Atlas-2.2.el6.x86_64.rpm
使用RPM -hiv Atlas-2.2.el6.x86_64.rpm
安装成功
配置文件：
[mysql-proxy]
#带#号的为非必需的配置项目
#管理接口的用户名
admin-username=admin
#管理接口的密码
admin-password=123
#Atlas后端连接的MySQL主库的IP和端口，可设置多项，用逗号分隔
proxy-backend-addresses=192.168.0.42:3306
#Atlas后端连接的MySQL从库的IP和端口，@后面的数字代表权重，用来作负载均衡，若省略则默认为1，可设置多项，用逗号分隔
proxy-read-only-backend-addresses=192.168.0.43:3306,192.168.0.44:3306
#用户名与其对应的加密过的MySQL密码，密码使用PREFIX/bin目录下的加密程序encrypt加密，下行的user1和user2为示例，将其替换为你的MySQL的用户名和加密密码！
pwds=atlas:3yb5jEku5h4=
#设置Atlas的运行方式，设为true时为守护进程方式，设为false时为前台方式，一般开发调试时设为false，线上运行时设为true,true后面不能有空格。
daemon=true
#设置Atlas的运行方式，设为true时Atlas会启动两个进程，一个为monitor，一个为worker，monitor在worker意外退出后会自动将其重启，设为false时只有worker，没有monitor，一般开发调试时设为false，线上运
行时设为true,true后面不能有空格。
keepalive=true
#工作线程数，对Atlas的性能有很大影响，可根据情况适当设置
event-threads=4
#日志级别，分为message、warning、critical、error、debug五个级别
log-level=message
#日志存放的路径
log-path=/usr/local/mysql-proxy/log
#SQL日志的开关，可设置为OFF、ON、REALTIME，OFF代表不记录SQL日志，ON代表记录SQL日志，REALTIME代表记录SQL日志且实时写入磁盘，默认为OFF
sql-log=ON
#慢日志输出设置。当设置了该参数时，则日志只输出执行时间超过sql-log-slow（单位：ms)的日志记录。不设置该参数则输出全部日志。
#sql-log-slow = 10
#实例名称，用于同一台机器上多个Atlas实例间的区分
instance=test
#Atlas监听的工作接口IP和端口
proxy-address=0.0.0.0:1234
#Atlas监听的管理接口IP和端口
admin-address=0.0.0.0:2345
#分表设置，此例中person为库名，mt为表名，id为分表字段，3为子表数量，可设置多项，以逗号分隔，若不分表则不需要设置该项
#tables = person.mt.id.3
#默认字符集，设置该项后客户端不再需要执行SET NAMES语句
charset = utf8
#允许连接Atlas的客户端的IP，可以是精确IP，也可以是IP段，以逗号分隔，若不设置该项则允许所有IP连接，否则只允许列表中的IP连接
client-ips=127.0.0.1,0.192.168.0
#Atlas前面挂接的LVS的物理网卡的IP(注意不是虚IP)，若有LVS且设置了client-ips则此项必须设置，否则可以不设置
#lvs-ips = 192.168.1.1
========================华丽的分割线============================
根据上面的配置文件，得到账户信息如下：
Atlas管理账号 ：admin
Atlas管理账号密码：123
Atlas管理账号端口:2345
Atlas工作账号：atlas
Atlas工作账号密码：atlas:3yb5jEku5h4= (123,用encrypt编译过的)
Atlas工作账号端口:1234
在主库及从库机器进行账号授权：
mysql>GRANT all privileges ON *.* TO ‘atlas‘@‘192.168.0.%‘ IDENTIFIED BY ‘123‘;
3台机器全部如此授权，在Atlas服务器上远程登陆3台机器成功.
在Atlas服务器验证Atlas管理账号及工作账号：
[[email protected] bin]# /usr/local/mysql-proxy/bin/mysql-proxyd test start
OK: MySQL-Proxy of test is started
[[email protected] bin]# ps -ef | grep mysql | grep -v grep
root 6638 5610 0 02:00 pts/2 00:00:00 mysql -h127.0.0.1 -P1234 -uatlas -px x
root 6667 1919 0 02:01 pts/1 00:00:00 mysql -h127.0.0.1 -P2345 -uadmin -px x
root 6695 1 0 02:06 ? 00:00:00 /usr/local/mysql-proxy/bin/mysql-proxy --defaults-file=/usr/local/mysql-proxy/conf/test.cnf
root 6696 6695 0 02:06 ? 00:00:00 /usr/local/mysql-proxy/bin/mysql-proxy --defaults-file=/usr/local/mysql-proxy/conf/test.cnf
[[email protected] ~]# mysql -h127.0.0.1 -P2345 -uadmin -p123
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 1
Server version: 5.0.99-agent-admin
Copyright (c) 2000, 2011, 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.
mysql> select * from help;
+-------------------+---------------------------------------+
| command | description |
+-------------------+---------------------------------------+
| SELECT * FROM help | shows this help |
| SELECT * FROM backends | lists the backends and their state |
| SET OFFLINE $backend_id | offline backend server, $backend_id is backend_ndx‘s id |
| SET ONLINE $backend_id | online backend server, ... |
| ADD MASTER $backend | example: "add master 127.0.0.1:3306", ... |
| ADD SLAVE $backend | example: "add slave 127.0.0.1:3306", ... |
| REMOVE BACKEND $backend_id | example: "remove backend 1", ... |
| ADD CLIENT $client | example: "add client 192.168.1.2", ... |
| REMOVE CLIENT $client | example: "remove client 192.168.1.2", ... |
| SAVE CONFIG | save the backends to config file |
+-------------------+---------------------------------------+
10 rows in set (0.00 sec)
mysql> select * from backends;
+-------------+-------------------+-------+------+
| backend_ndx | address | state | type |
+-------------+-------------------+-------+------+
| 1 | 192.168.0.42:3306 | down | rw |
| 2 | 192.168.0.44:3306 | down | ro |
| 3 | 192.168.0.43:3306 | down | ro |
+-------------+-------------------+-------+------+
3 rows in set (0.00 sec)
噩梦就此开始
登陆atlas工作账号：
[[email protected] ~]# mysql -h127.0.0.1 -P1234 -uatlas -p123
Welcome to the MySQL monitor. Commands end with ; or \g.
Your MySQL connection id is 1
Server version: 5.0.81-log
Copyright (c) 2000, 2011, 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.
mysql> show databases;
ERROR 2006 (HY000): MySQL server has gone away
No connection. Trying to reconnect...
Connection id: 2
Current database: *** NONE ***
ERROR 2013 (HY000): Lost connection to MySQL server during query
mysql>
无法连接到服务器
在管理账号中，将所有机器上线
mysql> SET ONLINE 1;
+-------------+-------------------+---------+------+
| backend_ndx | address | state | type |
+-------------+-------------------+---------+------+
| 1 | 192.168.0.42:3306 | unknown | rw |
+-------------+-------------------+---------+------+
1 row in set (0.00 sec)
再查看backends状态：
mysql> select * from backends;
+-------------+-------------------+-------+------+
| backend_ndx | address | state | type |
+-------------+-------------------+-------+------+
| 1 | 192.168.0.42:3306 | down | rw |
| 2 | 192.168.0.44:3306 | down | ro |
| 3 | 192.168.0.43:3306 | down | ro |
+-------------+-------------------+-------+------+
3 rows in set (0.00 sec)
依然如此
参照网上的方法，在3台主从机上的my.cnf中加入以下参数：
max_connect_errors=100000
max_allowed_packet=16777216 (16M)
依然不行
mysql> select * from backends;
+-------------+-------------------+-------+------+
| backend_ndx | address | state | type |
+-------------+-------------------+-------+------+
| 1 | 192.168.0.42:3306 | down | rw |
| 2 | 192.168.0.44:3306 | down | ro |
| 3 | 192.168.0.43:3306 | down | ro |
+-------------+-------------------+-------+------+
3 rows in set (0.00 sec)
咨询开发人员，得知：
client-ips=127.0.0.1,0.192.168.0
此选项，如非必须，不要打开
故障排除后，进行读写分离测试
在atlas客户端，创建数据库
mysql> create database atlas；
Query OK, 1 row affected (10.02 sec)
mysql> create table atlas1
-> (
-> name varchar(15) not null,
-> sex enum(‘M‘,‘F‘) not null,
-> age int(2) not null,
-> primary key(‘name‘)
-> );
mysql> desc atlas.atlas1;
+-------+---------------+------+-----+---------+-------+
| Field | Type | Null | Key | Default | Extra |
+-------+---------------+------+-----+---------+-------+
| name | varchar(15) | NO | PRI | NULL | |
| sex | enum(‘M‘,‘F‘) | NO | | NULL | |
| age | int(2) | NO | | NULL | |
+-------+---------------+------+-----+---------+-------+
3 rows in set (0.01 sec)
插入数据：
mysql> insert into atlas.atlas1
-> VALUES
-> (‘Zhangxuan‘,‘M‘,‘30‘),
-> (‘Yangyanjie‘,‘M‘,‘28‘),
-> (‘Liuchang‘,‘F‘,‘22‘);
Query OK, 3 rows affected (0.01 sec)
Records: 3 Duplicates: 0 Warnings: 0
在主库上进行一致性检查：
[[email protected] ~]# pt-table-checksum --user=root --host=192.168.0.42 --nocheck-replication-filters --nocheck-binlog-format
TS ERRORS DIFFS ROWS CHUNKS SKIPPED TIME TABLE
10-28T04:36:06 0 0 3 1 0 0.028 atlas.atlas1
10-28T04:36:06 0 0 51 1 0 0.027 mhatest.mha
10-28T04:36:06 0 0 3012 1 0 0.282 mhatest.num1
10-28T04:36:06 0 0 5 1 0 0.029 mhatest.test
成功，然后在从库1（192.168.0.43）和从库2（192.168.0.43）上各增加一条数据
从库1（192.168.0.43）：
mysql> INSERT INTO atlas.atlas1 VALUES (‘Yangzhenyu‘,‘M‘,‘30‘);
Query OK, 1 row affected (0.01 sec)
mysql> select * from atlas1;
+--------+---+--+
| name | sex | age |
+--------+---+--+
| Liuchang | F | 22 |
| Yangyanjie | M | 28 |
| Yangzhenyu | M | 30 |
| Zhangxuan | M | 30 |
+--------+---+---+
4 rows in set (0.00 sec)
从库2（192.168.0.44）
mysql> INSERT INTO atlas.atlas1 VALUES (‘Chenfei‘,‘M‘,‘14‘)；
mysql> select * from atlas.atlas1;
+--------+---+--+
| name | sex | age |
+--------+---+---+
| Chenfei | M | 14 |
| Liuchang | F | 22 |
| Yangyanjie | M | 28 |
| Zhangxuan | M | 30 |
+--------+---+---+
4 rows in set (0.00 sec)
此时，主库（192.168.0.42）上：
mysql> select * from atlas.atlas1;
+--------+---+---+
| name | sex | age |
+--------+---+---+
| Liuchang | F | 22 |
| Yangyanjie | M | 28 |
| Zhangxuan | M | 30 |
+--------+---+---+
3 rows in set (0.00 sec)
开始验证读写分离：
在atlas服务器上执行：
mysql> select * from atlas.atlas1 where name=‘Yangzhenyu‘;
Empty set (0.00 sec) /*第一次执行居然无返回值*/
mysql> select * from atlas.atlas1 where name=‘Yangzhenyu‘;
+------------+-----+-----+
| name | sex | age |
+------------+-----+-----+
| Yangzhenyu | M | 30 |
+------------+-----+-----+
1 row in set (0.00 sec) /*第二次正确返回*/
查询日志：
[10/28/2014 03:10:07] C:127.0.0.1:39798 S:192.168.0.43:3306 OK 0.719 "select * from atlas.atlas1 where name=‘Yangzhenyu‘"
[10/28/2014 03:10:10] C:127.0.0.1:39798 S:192.168.0.44:3306 OK 0.921 "select * from atlas.atlas1 where name=‘Yangzhenyu‘"
第一次无返回值原因：两个read-only库数据不一致导致，将两个从库数据一致后，查询
返回值正常：
mysql> select * from atlas.atlas1 where name=‘Yangzhenyu‘;
+--------+---+---+
| name | sex | age |
+--------+---+---+
| Yangzhenyu | M | 30 |
+--------+---+---+
1 row in set (0.00 sec) /*第一次查询*/
mysql> select * from atlas.atlas1 where name=‘Yangzhenyu‘;
+--------+---+---+
| name | sex | age |
+--------+---+---+
| Yangzhenyu | M | 30 |
+--------+---+---+
1 row in set (0.00 sec) /*第二次查询*/
读写分离成功
时间有限，下一步准备用sysbench进行压力测试