前言
为了降低一台数据库的IO，远程连接数据库的时候，可以实现读写分离进行调度。这里就出现了一个单点，所以必须要做一个高可用。当然数据库服务器也需要做主从复制。
实验结构
说明：上图的拓扑只是整个架构中的一个小部分，其余功能的实现此图并未规划出来。此拓扑实现的目的是利用proxysql实现数据读写分离，并对proxysql高可用。两台安装了Keepalived和proxysql虚拟成一个VIP对外提供服务。这两台mysql服务器做的半同步复制，192.168.32.111是主节点负责用户的写操作，192.168.32.112从节点负责用户的读操作。
实验步骤
1、安装数据库并配置半同步复制
主mysql：192.168.32.111
]#yum install mariadb -y <===centos6为mysql ]#vim /etc/my.cnf.d/server.cnf [server] skip_name_resolve = ON innodb_file_per_table = ON max_connection = 2000 log_bin = master-log server_id = 1 ]#systemctl start mariadb ]#mysql MariaDB [(none)]> GRANT REPLICATION SLAVE,REPLICATION CLIENT ON *.* TO 'ma'@'192.168.32.%' IDENTIFIED BY 'centos'; <===授权从节点复制的帐号 MariaDB [(none)]> SHOW MASTER LOGS; <===查看当前二进制日志文件 +-------------------+-----------+ | Log_name |File_size | +-------------------+-----------+ | master-log.000001 | 490 | +-------------------+-----------+ MariaDB [(none)]> SHOW BINLOG EVENTS IN 'master-log.000001'; <===查看日志内容保存点，从节点可以根据需要选择从哪开始复制，我们这里是测试环境，数据库为空，先从最开始复制。 +-------------------+-----+-------------+-----------+-------------+-------------------------------------------------------------------------------------------------+ | Log_name | Pos |Event_type | Server_id | End_log_pos |Info | +-------------------+-----+-------------+-----------+-------------+-------------------------------------------------------------------------------------------------+ | master-log.000001 | 4 | Format_desc | 1 | 245 | Server ver: 5.5.52-MariaDB, Binlog ver: 4 | | master-log.000001 | 245 | Query | 1 | 415 | grant replication slave,replication client on *.* to'ma'@'192.168.32.%' identified by 'centos' | | master-log.000001 | 415 | Query | 1 | 490 | flush privileges | +-------------------+-----+-------------+-----------+-------------+-------------------------------------------------------------------------------------------------+ MariaDB [(none)]> INSTALL PLUGIN rpl_semi_sync_master SONAME 'semisync_master'; <===安装插件 MariaDB [(none)]> SHOW GLOBALVARIABLES LIKE '%rpl%'; <===安装完以后不代表开启。 +------------------------------------+-------+ | Variable_name | Value | +------------------------------------+-------+ | rpl_recovery_rank | 0 | | rpl_semi_sync_master_enabled | OFF | | rpl_semi_sync_master_timeout | 10000 | | rpl_semi_sync_master_trace_level | 32 | | rpl_semi_sync_master_wait_no_slave | ON | +------------------------------------+-------+ MariaDB [(none)]> SHOW GLOBAL STATUS LIKE '%rpl%'; <===查看复制各种状态 MariaDB [(none)]> SET @@GLOBAL.rpl_semi_sync_master_enabled=ON; <===开启插件 MariaDB [(none)]> FLUSH PRIVILEGES;
从mysql配置：192.168.32.112
]#yum install mariadb -y <===centos6为mysql ]#vim /etc/my.cnf.d/server.cnf [server] skip_name_resolve = ON innodb_file_per_table = ON max_connection = 2000 relay_log = relay-log server_id = 2 ]#systemctl start mariadb ]#mysql > CHANGE MASTER TO MASTER_HOST='192.168.32.111',MASTER_USER='ma',MASTER_PASSWORD='centos',MASTER_LOG_FILE='master-log.000001',MASTER_LOG_POS=245; >START SLAVE; > SHOW SLAVE STATUS\G; *************************** 1. row *************************** Slave_IO_State: Waiting for master to send event Master_Host: 192.168.32.111 <===主节点 Master_User: ma <===用户 Master_Port: 3306 <===端口 Connect_Retry: 60 Master_Log_File: master-log.000001 <===从哪个文件复制 Read_Master_Log_Pos: 490 <===哪个位置 Relay_Log_File: relay-log.000002 <===复制到本地的中继日志 Relay_Log_Pos: 775 <===中继日志的位置 Relay_Master_Log_File: master-log.000001 Slave_IO_Running: Yes <===复制要靠IO线程 Slave_SQL_Running: Yes <===重放要靠SQL线程 … > INSTALL PLUGIN rpl_semi_sync_slave SONAME 'semisync_slave'; > SET @@GLOBAL.rpl_semi_sync_slave_enabled=ON; > SHOW GLOBAL VARIABLES LIKE '%rpl%'; <===查看变量里已经开启了 +---------------------------------+-------+ | Variable_name | Value | +---------------------------------+-------+ | rpl_recovery_rank | 0 | | rpl_semi_sync_slave_enabled | ON | | rpl_semi_sync_slave_trace_level | 32 | +---------------------------------+-------+ > SHOW GLOBAL STATUS LIKE '%rpl%'; <===但是，这个时候看状态依然是关闭的！ +----------------------------+-------------+ | Variable_name | Value | +----------------------------+-------------+ | Rpl_semi_sync_slave_status | OFF | | Rpl_status | AUTH_MASTER | +----------------------------+-------------+ > STOP SLAVE IO_THREAD; > START SLAVEIO_THREAD; <===需要重启下复制线程
总结：到这里我们已经实现了数据库的半同步复制，如果在主节点上创建数据从节点能同步到，证明实验第一步成功了。为了后面读写分离可以看到效果，可以考虑在从节点上加个过滤器，某一个数据库中数据不
同步到从节点上。过滤机制有两种方法：1、主节点二进制日志中进行过滤；2、在从节点的中继日志中进行过滤。这里我们使用第二种方法。
SET @@GLOBAL.replicate_ignore_DB=hidb; <===在从节点上设置过滤hidb的库
2、配置proxysql：192.168.32.104
]#yum install -y \.proxysql-1.4.2-1-centos7.x86_64.rpm <===我是下载好上传到本地的，安装包见附件。不排除有依赖关系，所以用yum安装。
]#rpm -ql proxysql <===包组很少
/etc/init.d/proxysql
/etc/proxysql.cnf
/usr/bin/proxysql
/usr/share/proxysql/tools/proxysql_galera_checker.sh
/usr/share/proxysql/tools/proxysql_galera_writer.pl
]# vim /etc/proxysql.cnf
admin_variables= <===管理接口的用户和账户：mysql -S /tmp/proxysql_admin.sock -uadmin -padmin.安全起见，建议更改
{
admin_credentials="admin:admin"
mysql_ifaces="127.0.0.1:6032;/tmp/proxysql_admin.sock"
}
mysql_variables=
{
threads=4
max_connections=2048
default_query_delay=0
default_query_timeout=36000000
have_compress=true
poll_timeout=2000
interfaces="0.0.0.0:3306;/tmp/mysql.sock"
default_schema="information_schema"
stacksize=1048576
server_version="5.5.30"
connect_timeout_server=3000
monitor_username="monitor"
monitor_password="monitor"
monitor_history=600000
monitor_connect_interval=60000
monitor_ping_interval=10000
monitor_read_only_interval=1500
monitor_read_only_timeout=500
ping_interval_server_msec=120000
ping_timeout_server=500
commands_stats=true
sessions_sort=true
connect_retries_on_failure=10
}
mysql_servers = <===定义两个mysql服务器地址、端口、属于哪个组（下面有定义）、是否压缩、权重
(
{
address = "192.168.32.111"
port = 3306
hostgroup = 0
status = "ONLINE"
weight = 1
compression = 0
},
{
address = "192.168.32.112"
port = 3306
hostgroup = 1
status = "ONLINE"
weight = 1
compression = 0
}
)
mysql_users: <===连接数据库的用户和账户，和之前授权复制主从的一致
(
{
username = "dbadmin"
password = "centos"
default_hostgroup = 0
active = 1
}
)
mysql_query_rules: <===数据查询规则定义
(
{
rule_id=1
active=1
match_pattern="^SELECT .* FOR UPDATE$"
destination_hostgroup=0
apply=1
},
{
rule_id=2
active=1
match_pattern="^SELECT"
destination_hostgroup=1
apply=1
}
)
mysql_replication_hostgroups=
(
{
writer_hostgroup=0 <===0组为写
reader_hostgroup=1 <===1组为读
comment="test repl 1" <===描述信息
}
)
]#service proxysql start
]# mysql -udbadmin -pcentos -h172.18.32.104 <===通过本地访问数据库
>INSERT INTO hidb.tal1 VALUES (4,'li'); <===插入一个数据,写入主节点
>SELECT * FROM hidb.tal1; <===读数据是从从点读，而从节点做了hidb的过滤，因此是读不到插入的这一行数据。实现了读写分离。
总结：到这里我们是已经实现了数据库的读写分离。下面我们还需要对proxysql做高可用。
3、在另外一台主机192.168.32.105上配置proxysql，具体配置和上面相同。
4、电脑配置有限，这里就在proxysql服务器上配置Keepalived。
keepalived：192.168.32.104
]# yum install -y keepalived
]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
[email protected]
[email protected]
[email protected]
}
notification_email_from [email protected]
smtp_server 192.168.200.1
smtp_connect_timeout 30
router_id node1 <===添加
vrrp_mcast_group4 224.100.100.100 <===组播地址
}
vrrp_script chk_nginx { <===检测服务的脚本
script "killall -0 proxysql &> /dev/null && exit 0 || exit 1"
interval 1
weight -30
fall 3
rise 3
}
vrrp_instance VI_1 {
state MASTER
interface ens33
virtual_router_id 22
priority 100
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
192.168.32.99/24 <===虚拟一个VIP对外提供服务
}
track_script { <===调用上面的脚本
chk_nginx
}
}
]#systemctl start keepalived
keepalived：192.168.32.104
]# yum install -y keepalived
]# vim /etc/keepalived/keepalived.conf
! Configuration File for keepalived
global_defs {
notification_email {
[email protected]
[email protected]
[email protected]
}
notification_email_from [email protected]
smtp_server 192.168.200.1
smtp_connect_timeout 30
router_id node2
vrrp_mcast_group4 224.100.100.100
}
vrrp_script chk_nginx {
script "killall -0 proxysql &> /dev/null && exit 0 || exit 1"
interval 1
weight -10
fall 3
rise 3
}
vrrp_instance VI_1 {
state backup
interface ens33
virtual_router_id 22
priority 90
advert_int 1
authentication {
auth_type PASS
auth_pass 1111
}
virtual_ipaddress {
192.168.32.99/24
}
track_script {
chk_nginx
}
}
]#systemctl start keepalived
5、测试
]#tcpdump -i ens33 -nn host 224.100.100.100 <===对广播地址进行抓包
在客户端使用VIP使用，实现效果：当主proxysql服务down掉以后，从proxysql服务器拿到VIP开始提供服务。实现了高可用
]# mysql -udbadmin -pcentos -h192.168.32.99
6、补充
我们可以在proxysql服务器上使用管理接口登录上去看看状态
]# mysql -S /tmp/proxysql_admin.sock -uadmin -padmin > SHOW DATABASES; +-----+---------+-------------------------------+ | seq | name | file | +-----+---------+-------------------------------+ | 0 | main | | | 2 | disk | /var/lib/proxysql/proxysql.db | | 3 | stats | | | 4 | monitor | | +-----+---------+-------------------------------+ > SELECT * FROM mysql_servers; +--------------+----------------+------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | hostgroup_id | hostname | port | status | weight | compression | max_connections | max_replication_lag | use_ssl | max_latency_ms | comment | +--------------+----------------+------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+ | 0 | 192.168.32.111 | 3306 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | | 1 | 192.168.32.112 | 3306 | ONLINE | 1 | 0 | 1000 | 0 | 0 | 0 | | +--------------+----------------+------+--------+--------+-------------+-----------------+---------------------+---------+----------------+---------+
总结：还有很多表就不一一展示了。这里通过proxysql+keepalived实现了数据库简单的读写分离，简单的高可用。实际应用中比这复杂的多，并不是一个人能完成的。