wKiom1gcMPnQK8fWAAA4eVx2Dz8985.jpg

wKioL1gcMPnydlzeAAA4pG6yXEQ878.jpg

MySQL读写分离与负载均衡--MHA与MaxScale

环境介绍

Part1:写在最前

看了某大牛的文章，讲述了一下MaxScale比LVS的好处多多，那您倒是放出来配置文件啊~~大牛说：

需要的单独找我吧，太长了配置文件……

看到这我心中久久不能平静啊。。。联系不上您呐 = =，于是各种资料各种找啊~各种坑各种血崩啊！~~~

由于不知道大牛的配置文件是什么样子，本文仅以随笔的形式，记录下实施过程。也欢迎您和我探讨您在实施MaxScale时遇到的各种问题和心得体会。


Part2:环境

MySQL5.7 MHA + MaxScale2.0

192.168.1.248 HE1 slave1

192.168.1.249 HE2 slave2

192.168.1.250 HE3 master

192.168.1.251 HE4 MHA-manager

192.168.1.100 MHA-vip


Part3:MHA

MHA的优点不作赘述，看下原理图吧

MySQL5.7MHA+MaxScale2.0构建高可用环境_第1张图片

从宕机崩溃的Master保存二进制日志事件(binlogevent)

识别含有最新更新的Slave

应用差异的中继日志(relaylog)到其他Slave

应用从Master保存的二进制日志事件

提升一个Slave为新的Master

使其他的Slave连接新的Master进行复制



构建MySQL5.7MHA

Part1:写在最前

MHA的部署不是本文的叙述重点，网上比比皆是。这里只记录下MySQL5.7的MHA搭建时的一些坑


Part2:坑

①mha4mysql-manager-0.57.tar.gz

②mha4mysql-node-0.57.tar.gz

Warning:警告这两个包首先你要搞到，虽然说0.56什么的不代表支持的mysql版本，但经过测试，想要在MySQL5.7上部署MHA，少走坑，请用0.57的。


Part3:安装包的位置

请在所有的节点包括Manager节点安装好你的mha4mysql-node-0.57.tar.gz，以保证后期在管理节点执行perl Makefile.PL的时候，你能如愿以偿的看到如下信息

[root@HE4 mha4mysql-manager-0.57]# perl Makefile.PL 
*** Module::AutoInstall version 1.06
*** Checking for Perl dependencies...
[Core Features]
- DBI                   ...loaded. (1.609)
- DBD::mysql            ...loaded. (4.013)
- Time::HiRes           ...loaded. (1.9721)
- Config::Tiny          ...loaded. (2.12)
- Log::Dispatch         ...loaded. (2.26)
- Parallel::ForkManager ...loaded. (0.7.5)
- MHA::NodeConst        ...loaded. (0.57)
*** Module::AutoInstall configuration finished.
Writing Makefile for mha4mysql::manager


Part4:一些常见错误记录

如果遇到

①这样

[root@HE2 bin]# masterha_check_repl --conf=/etc/mha/mha.conf 
Tue Apr  5 22:09:32 2016 - [warning] Global configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Apr  5 22:09:32 2016 - [info] Reading application default configuration from /etc/mha/mha.conf..
Tue Apr  5 22:09:32 2016 - [info] Reading server configuration from /etc/mha/mha.conf..
Tue Apr  5 22:09:32 2016 - [info] MHA::MasterMonitor version 0.57.
Tue Apr  5 22:09:32 2016 - [error][/usr/local/lib64/perl5/MHA/ServerManager.pm, ln188] There is no alive server. We can't do failover
Tue Apr  5 22:09:32 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln424] Error happened on checking configurations.  at /usr/local/lib64/perl5/MHA/MasterMonitor.pm line 326
Tue Apr  5 22:09:32 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln523] Error happened on monitoring servers.
Tue Apr  5 22:09:32 2016 - [info] Got exit code 1 (Not master dead).

解决方案

用的不是默认端口3306，请修改你的配置文件


②这样

[root@HE4 ~]#
masterha_check_repl --conf=/etc/mha/mha.conf
Tue Apr  5 22:36:33 2016 - [warning] Global
configuration file /etc/masterha_default.cnf not found. Skipping.
Tue Apr  5 22:36:33 2016 - [info] Reading application
default configuration from /etc/mha/mha.conf..
Tue Apr  5 22:36:33 2016 - [info] Reading server
configuration from /etc/mha/mha.conf..
Tue Apr  5 22:36:33 2016 - [info] MHA::MasterMonitor
version 0.57.
Tue Apr  5 22:36:34 2016 - [info] GTID failover mode =
0
Tue Apr  5 22:36:34 2016 - [info] Dead Servers:
Tue Apr  5 22:36:34 2016 - [info] Alive Servers:
Tue Apr  5 22:36:34 2016 - [info]   192.168.1.250(192.168.1.250:4008)
Tue Apr  5 22:36:34 2016 - [info]   192.168.1.248(192.168.1.248:4008)
Tue Apr  5 22:36:34 2016 - [info]   192.168.1.249(192.168.1.249:4008)
Tue Apr  5 22:36:34 2016 - [info] Alive Slaves:
Tue Apr  5 22:36:34 2016 - [info]   192.168.1.248(192.168.1.248:4008)  Version=5.6.16-log (oldest major version
between slaves) log-bin:enabled
Tue Apr  5 22:36:34 2016 - [info]     Replicating from
192.168.1.250(192.168.1.250:4008)
Tue Apr  5 22:36:34 2016 - [info]     Primary candidate for the new Master
(candidate_master is set)
Tue Apr  5 22:36:34 2016 - [info]   192.168.1.249(192.168.1.249:4008)  Version=5.6.16-log (oldest major version
between slaves) log-bin:enabled
Tue Apr  5 22:36:34 2016 - [info]     Replicating from
192.168.1.250(192.168.1.250:4008)
Tue Apr  5 22:36:34 2016 - [info]     Not candidate for the new Master
(no_master is set)
Tue Apr  5 22:36:34 2016 - [info] Current Alive
Master: 192.168.1.250(192.168.1.250:4008)
Tue Apr  5 22:36:34 2016 - [info] Checking slave
configurations..
Tue Apr  5 22:36:34 2016 - [warning]  relay_log_purge=0 is not set on slave
192.168.1.248(192.168.1.248:4008).
Tue Apr  5 22:36:34 2016 - [warning]  relay_log_purge=0 is not set on slave
192.168.1.249(192.168.1.249:4008).
Tue Apr  5 22:36:34 2016 - [info] Checking replication
filtering settings..
Tue Apr  5 22:36:34 2016 - [info]  binlog_do_db= , binlog_ignore_db=
Tue Apr  5 22:36:34 2016 - [info]  Replication filtering check ok.
Tue Apr  5 22:36:34 2016 - [info] GTID (with auto-pos)
is not supported
Tue Apr  5 22:36:34 2016 - [info] Starting SSH
connection tests..
Tue Apr  5 22:36:35 2016 - [info] All SSH connection
tests passed successfully.
Tue Apr  5 22:36:35 2016 - [info] Checking MHA Node
version..
Tue Apr  5 22:36:36 2016 - [info]  Version check ok.
Tue Apr  5 22:36:36 2016 - [info] Checking SSH
publickey authentication settings on the current master..
Tue Apr  5 22:36:36 2016 - [info] HealthCheck: SSH to
192.168.1.250 is reachable.
Tue Apr  5 22:36:36 2016 - [info] Master MHA Node
version is 0.57.
Tue Apr  5 22:36:36 2016 - [info] Checking recovery
script configurations on 192.168.1.250(192.168.1.250:4008)..
Tue Apr  5 22:36:36 2016 - [info]   Executing command: save_binary_logs
--command=test --start_pos=4 --binlog_dir=/log/mysql
--output_file=/usr/local/mha/save_binary_logs_test --manager_version=0.57
--start_file=mysql-bin.000009
Tue Apr  5 22:36:36 2016 - [info]   Connecting to
[email protected](192.168.1.250:22)..
  Creating /usr/local/mha if not exists..
Creating directory /usr/local/mha.. done.
   ok.
  Checking output directory is accessible or
not..
   ok.
  Binlog found at /log/mysql, up to
mysql-bin.000009
Tue Apr  5 22:36:36 2016 - [info] Binlog setting check
done.
Tue Apr  5 22:36:36 2016 - [info] Checking SSH
publickey authentication and checking recovery script configurations on all
alive slave servers..
Tue Apr  5 22:36:36 2016 - [info]   Executing command : apply_diff_relay_logs
--command=test --slave_user='root' --slave_host=192.168.1.248
--slave_ip=192.168.1.248 --slave_port=4008 --workdir=/usr/local/mha
--target_version=5.6.16-log --manager_version=0.57
--relay_log_info=/data/mysql/relay-log.info 
--relay_dir=/data/mysql/ 
--slave_pass=xxx
Tue Apr  5 22:36:36 2016 - [info]   Connecting to
[email protected](192.168.1.248:22)..
Can't exec
"mysqlbinlog": No such file or directory at
/usr/local/lib64/perl5/MHA/BinlogManager.pm line 106.
mysqlbinlog version
command failed with rc 1:0, please verify PATH, LD_LIBRARY_PATH, and client
options
 at /usr/local/bin/apply_diff_relay_logs line
493
Tue Apr  5 22:36:36 2016 -
[error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln205] Slaves settings
check failed!
Tue Apr  5 22:36:36 2016 -
[error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln413] Slave configuration
failed.
Tue Apr  5 22:36:36 2016 -
[error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln424] Error happened on
checking configurations.  at
/usr/local/bin/masterha_check_repl line 48
Tue Apr  5 22:36:36 2016 -
[error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln523] Error happened on
monitoring servers.
Tue Apr  5 22:36:36 2016 - [info] Got exit code 1 (Not
master dead).
 
MySQL Replication
Health is NOT OK!


解决方案

[root@HE1 MHA]# ln -s /usr/local/mysql/bin/mysqlbinlog /usr/bin/mysqlbinlog



③或者这样

  Binlog found at /log/mysql, up to mysql-bin.000009
Tue Apr  5 22:43:55 2016 - [info] Binlog setting check done.
Tue Apr  5 22:43:55 2016 - [info] Checking SSH publickey authentication and checking recovery script configurations on all alive slave servers..
Tue Apr  5 22:43:55 2016 - [info]   Executing command : apply_diff_relay_logs --command=test --slave_user='root' --slave_host=192.168.1.248 --slave_ip=192.168.1.248 --slave_port=4008 --workdir=/usr/local/mha --target_version=5.6.16-log --manager_version=0.57 --relay_log_info=/data/mysql/relay-log.info  --relay_dir=/data/mysql/  --slave_pass=xxx
Tue Apr  5 22:43:55 2016 - [info]   Connecting to [email protected](192.168.1.248:22).. 
mysqlbinlog: unknown variable 'default-character-set=utf8'
mysqlbinlog version command failed with rc 7:0, please verify PATH, LD_LIBRARY_PATH, and client options
 at /usr/local/bin/apply_diff_relay_logs line 493
Tue Apr  5 22:43:55 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln205] Slaves settings check failed!
Tue Apr  5 22:43:55 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln413] Slave configuration failed.
Tue Apr  5 22:43:55 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln424] Error happened on checking configurations.  at /usr/local/bin/masterha_check_repl line 48
Tue Apr  5 22:43:55 2016 - [error][/usr/local/lib64/perl5/MHA/MasterMonitor.pm, ln523] Error happened on monitoring servers.
Tue Apr  5 22:43:55 2016 - [info] Got exit code 1 (Not master dead).

 

解决方案

注释掉my.cnf中的

[client]

#default-character-set=utf8


遇到上述错误别慌张，看日志，根据报错来排查问题。


折腾半天，就为了这个OK


[root@HE4 mha4mysql-manager-0.57]# masterha_check_status --conf=/etc/mha/mha.conf

mha (pid:32726) is running(0:PING_OK), master:192.168.1.250



MaxScale2.0

Part1:写在最前

maxscale是mariadb公司开发的一套数据库中间件，可以很方便的实现读写分离方案；并且提供了读写分离的负载均衡和高可用性保障。另外maxscale对于前端应用而言是透明的，我们可以很方便的将应用迁移到maxscale中实现读写分离方案，来分担主库的压力。maxscale也提供了sql语句的解析过滤功能。这里我们主要讲解maxscale的安装、配置以及注意事项。


MySQL5.7MHA+MaxScale2.0构建高可用环境_第2张图片

Part2:整体架构

MySQL5.7MHA+MaxScale2.0构建高可用环境_第3张图片



Part3:安装

Maxscale配置很简单

[root@HE3 MHA]# yum -y install maxscale-2.0.1-2.centos.6.x86_64.rpm (只在Maxscale上执行)

[root@HE3 ~]# cat /etc/maxscale.cnf
# MaxScale documentation on GitHub:
# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Documentation-Contents.md

# Global parameters
#
# Number of threads is autodetected, uncomment for manual configuration
# Complete list of configuration options:
# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Getting-Started/Configuration-Guide.md

[maxscale]
threads=auto

# Server definitions
#
# Set the address of the server to the network
# address of a MySQL server.
#

[server1]
type=server
address=192.168.1.248
port=3306
protocol=MySQLBackend
myweight=5

[server2]
type=server
address=192.168.1.249
port=3306
protocol=MySQLBackend
myweight=5

[server3]
type=server
address=192.168.1.100
port=3306
protocol=MySQLBackend

# Monitor for the servers
#
# This will keep MaxScale aware of the state of the servers.
# MySQL Monitor documentation:
# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Monitors/MySQL-Monitor.md

[MySQL Monitor]
type=monitor
module=mysqlmon
servers=server1,server2,server3
user=mysync
passwd=MANAGER
monitor_interval=10000

# Service definitions
#
# Service Definition for a read-only service and
# a read/write splitting service.
#

# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Routers/ReadConnRoute.md

#[Read-Only Service]                       ###只读服务
#type=service
#router=readconnroute
#servers=server1,server2,server3
#user=sys_admin
#passwd=MANAGER
#router_options=slave

# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Routers/ReadWriteSplit.md



[Read-Write Service]                       ####写服务
type=service
router=readwritesplit
enable_root_user=1
servers=server1,server2,server3
user=sys_admin
passwd=MANAGER
weightby=myweight
router_options=slave_selection_criteria=LEAST_GLOBAL_CONNECTIONS
max_slave_connections=1


# https://github.com/mariadb-corporation/MaxScale/blob/master/Documentation/Reference/MaxAdmin.md

[MaxAdmin Service]
type=service
router=cli


#[Read-Only Listener]
#type=listener
#service=Read-Only Service
#protocol=MySQLClient
#port=4008                                  ##读服务启动监听 端口4008

[Read-Write Listener]
type=listener
service=Read-Write Service
protocol=MySQLClient
port=4006                                  ####写服务启动监听 端口

[MaxAdmin Listener]
type=listener
service=MaxAdmin Service
protocol=maxscaled
port=6603                                  ###管理端口


[root@HE3 ~]# /etc/init.d/maxscale start

Starting MaxScale: maxscale (pid 28851) is running...      [  OK  ]

 

[root@HE3 ~]#  netstat -lntp |grep maxscale

tcp        0      0 0.0.0.0:6603                0.0.0.0:*                   LISTEN      29878/maxscale     

tcp        0      0 0.0.0.0:4006                0.0.0.0:*                   LISTEN      29878/maxscale     



[root@HE3 ~]# maxadmin -pmariadb list services

Services.

--------------------------+----------------------+--------+---------------

Service Name              | Router Module        | #Users | Total Sessions

--------------------------+----------------------+--------+---------------

Read-Only Service         | readconnroute        |      1 |     5

Read-Write Service        | readwritesplit       |      1 |    11

MaxAdmin Service          | cli                  |      2 |     3

--------------------------+----------------------+--------+---------------

 

[root@HE3 ~]# maxadmin -pmariadb list servers

Servers.

-------------------+-----------------+-------+-------------+--------------------

Server             | Address         | Port  | Connections | Status             

-------------------+-----------------+-------+-------------+--------------------

server1            | 192.168.1.248   |  3306 |           0 | Slave, Running

server2            | 192.168.1.249   |  3306 |           0 | Slave, Running

server3            | 192.168.1.250   |  3306 |           0 | Master, Running

-------------------+-----------------+-------+-------------+--------------------


至此，MHA+Maxscale的环境就完成了。



读写分离与负载均衡校验

Part1:读写分离

在Master构建相应的测试表

mysql> select * from helei;
+--------+
| a      |
+--------+
| HE3    |
+--------+
3 rows in set (0.00 sec)


在slave1插入数据HE1

mysql> select * from helei;
+------+
| a    |
+------+
| HE3  |
| HE1  |
+------+


在slave2插入数据HE2

mysql> select * from helei;
+------+
| a    |
+------+
| HE3  |
| HE2  |
+------+


现在链接4006读写分离端口，进行数据写入

[root@HE3 ~]#  mysql -h192.168.1.250 -P 4006 -usys_admin -pMANAGER maxscale -e"insert into helei values('写入');"
mysql: [Warning] Using a password on the command line interface can be insecure.
[root@HE3 ~]#  mysql -h192.168.1.250 -P 4008 -usys_admin -pMANAGER maxscale -e"select * from helei;"
mysql: [Warning] Using a password on the command line interface can be insecure.
+--------+
| a      |
+--------+
| HE3    |
| HE1    |
| 写入   |
+--------+

可以看到主库插入完毕后从库已经同步完成，这条查询完成在了HE1（slave1）上


Part2:负载均衡

我们配置的read比例为1:1

[root@HE3 ~]# maxadmin -pmariadb show service "Read-Write Service"
Service 0xef5570
Service:                             Read-Write Service
Router:                              readwritesplit (0x7ff5e8fa6ec0)
State:                               Started
Number of router sessions:           15
Current no. of router sessions:      0
Number of queries forwarded:          41
Number of queries forwarded to master:2 (4.88%)
Number of queries forwarded to slave: 39 (95.12%)
Number of queries forwarded to all:   0 (0.00%)
Connection distribution based on myweight server parameter.
Server               Target %    Connections  Operations
                               Global  Router
server1              50.0%     0       0       0
server2              50.0%     0       0       0
server3              100.0%     0       0       0
Started:                             Thu Nov  3 23:46:27 2016
Root user access:                    Enabled
Backend databases:
192.168.1.248:3306  Protocol: MySQLBackend
192.168.1.249:3306  Protocol: MySQLBackend
192.168.1.100:3306  Protocol: MySQLBackend
Routing weight parameter:            myweight
Users data:                          0xf09370
Total connections:                   16
Currently connected:                 1



[root@HE3 ~]# for i in `seq 1 10`; do mysql -h 192.168.1.250 -P 4006 -usys_admin -pMANAGER maxscale -e "select @@hostname; select sleep(10)" 2>/dev/null & done
[root@HE3 ~]# +------------+
| @@hostname |
+------------+
| HE1        |
+------------+
+------------+
| @@hostname |
+------------+
| HE1        |
+------------+
+------------+
| @@hostname |
+------------+
| HE1        |
+------------+
+------------+
| @@hostname |
+------------+
| HE1        |
+------------+
+------------+
| @@hostname |
+------------+
| HE1        |
+------------+
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+


OK!That's how it works!~



故障测试

Part1:stop slave故障

停止HE1的复制

[root@HE1 ~]# mysql
-uroot -p
Enter password:
Welcome to the MySQL
monitor.  Commands end with ; or \g.
Your MySQL
connection id is 74
Server version:
5.7.16-log MySQL Community Server (GPL)
 
Copyright (c) 2000,
2016, 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> stop
slave;
Query OK, 0 rows
affected (0.05 sec)
 
[root@HE3 ~]#
maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port 
| Connections | Status             
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   | 
3306 |           0 | Running
server2            | 192.168.1.249   | 
3306 |           0 | Slave,
Running
server3            | 192.168.1.250   | 
3306 |           0 | Master,
Running
-------------------+-----------------+-------+-------------+--------------------
 
[root@HE3 ~]#  mysql -h192.168.1.250 -P 4006 -usys_admin
-pMANAGER maxscale -e"select * from helei;"
mysql: [Warning]
Using a password on the command line interface can be insecure.
+--------+
| a      |
+--------+
| HE3    |
| HE2    |
| 写入  
|
+--------+
[root@HE3 ~]#  mysql -h192.168.1.250 -P 4006 -usys_admin
-pMANAGER maxscale -e"select * from helei;"
mysql: [Warning]
Using a password on the command line interface can be insecure.
+--------+
| a      |
+--------+
| HE3    |
| HE2    |
| 写入  
|
+--------+

 

可以看出，在slave1故障后，所有的读操作都进入了HE2(slave2)；


 

 

 

恢复HE1

 
mysql> start
slave;
Query OK, 0 rows
affected (0.00 sec)
 
 
[root@HE3 ~]#
maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port 
| Connections | Status             
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   | 
3306 |           0 | Slave,
Running
server2            | 192.168.1.249   | 
3306 |           0 | Slave,
Running
server3            | 192.168.1.250   | 
3306 |           0 | Master,
Running
-------------------+-----------------+-------+-------------+--------------------
 
验证
[root@HE3 ~]#  mysql -h192.168.1.250 -P 4006 -usys_admin
-pMANAGER maxscale -e"select * from helei;"
mysql: [Warning]
Using a password on the command line interface can be insecure.
+--------+
| a      |
+--------+
| HE3    |
| HE2    |
| 写入  
|
+--------+
[root@HE3 ~]#  mysql -h192.168.1.250 -P 4006 -usys_admin
-pMANAGER maxscale -e"select * from helei;"
mysql: [Warning]
Using a password on the command line interface can be insecure.
+--------+
| a      |
+--------+
| HE3    |
| HE1    |
| 写入  
|
+--------+

 

在HE1（slave1）恢复完成后，重新有了负载均衡。

 

 

 

 

Part2:mysql down故障

[root@HE1 ~]#
/etc/init.d/mysqld stop
Shutting down
MySQL..... SUCCESS!
 
停止HE1(slave)可以看到转发到了HE2上
 
[root@HE3 ~]#  maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port 
| Connections | Status             
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   | 
3306 |           0 | Down
server2            | 192.168.1.249   | 
3306 |           1 | Slave,
Running
server3            | 192.168.1.250   | 
3306 |           1 | Master,
Running
-------------------+-----------------+-------+-------------+--------------------
[root@HE3
~]# mysql -h192.168.1.250 -P 4006 -usys_admin -pMANAGER
mysql: [Warning]
Using a password on the command line interface can be insecure.
Welcome to the MySQL
monitor.  Commands end with ; or \g.
Your MySQL
connection id is 28948
Server version:
5.5.5-10.0.0 2.0.1-maxscale MySQL Community Server (GPL)
 
Copyright (c) 2000,
2016, 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
@@hostname;
+------------+
| @@hostname |
+------------+
| HE2        |
+------------+
1 row in set (0.00
sec)
 
 
停掉2台slave，观察maxscale的状态
[root@HE3 ~]#
maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port 
| Connections | Status             
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   | 
3306 |           0 | Running
server2            | 192.168.1.249   | 
3306 |           0 | Running
server3            | 192.168.1.250   | 
3306 |           0 | Master, Stale
Status, Running
-------------------+-----------------+-------+-------------+--------------------



Warning:警告这里我并没有在

[MySQL Monitor]中配置detect_stale_master=true

可以看出，在maxscale2.0中，已经默认从库都停掉，也不影响


Part3:master故障

[root@HE3 ~]# maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port  | Connections | Status              
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   |  3306 |           0 | Slave, Running
server2            | 192.168.1.249   |  3306 |           0 | Slave, Running
server3            | 192.168.1.100   |  3306 |           0 | Master, Running
-------------------+-----------------+-------+-------------+--------------------
[root@HE3 ~]# ps -ef|grep mysql
root     27709     1  0 Nov03 ?        00:00:00 /bin/sh /usr/local/mysql/bin/mysqld_safe --datadir=/data/mysql --pid-file=/data/mysql/HE3.pid
mysql    28415 27709  0 Nov03 ?        00:00:19 /usr/local/mysql/bin/mysqld --basedir=/usr/local/mysql --datadir=/data/mysql --plugin-dir=/usr/local/mysql/lib/plugin --user=mysql --log-error=/data/mysql/error.log --open-files-limit=8192 --pid-file=/data/mysql/HE3.pid --socket=/tmp/mysql.sock --port=3306
root     30794 28966  0 02:34 pts/1    00:00:00 grep mysql
[root@HE3 ~]# kill -9 28415 27709
[root@HE3 ~]# maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port  | Connections | Status              
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   |  3306 |           0 | Slave, Running
server2            | 192.168.1.249   |  3306 |           0 | Slave, Running
server3            | 192.168.1.100   |  3306 |           0 | Master, Running
-------------------+-----------------+-------+-------------+--------------------
[root@HE3 ~]# maxadmin -pmariadb list servers
Servers.
-------------------+-----------------+-------+-------------+--------------------
Server             | Address         | Port  | Connections | Status              
-------------------+-----------------+-------+-------------+--------------------
server1            | 192.168.1.248   |  3306 |           0 | Master, Running
server2            | 192.168.1.249   |  3306 |           0 | Slave, Running
server3            | 192.168.1.100   |  3306 |           0 | Running
-------------------+-----------------+-------+-------------+--------------------

MySQL5.7MHA+MaxScale2.0构建高可用环境_第4张图片

可以看出，kill master的mysql进程后，MHA将slave1拉成新的master,Maxsale也识别了这一状态。



Java Druid注意事项

------------以下文章取材于贺春旸技术博客-------------------

http://hcymysql.blog.51cto.com/5223301/1869414

现象：

程序会不定时的出现连接错误，问题bug的异常信息如下：

The last packet successfully received from the server was 116 milliseconds ago.  The last packet sent successfully to the server was 115 milliseconds ago.

java.sql.SQLException: No database selected

 

但通过客户端sqlyog/navicat连接均为正常。

----------------------------------------------------------------------

另：maxscale1.4.3有时还会出现挂起现象，4006端口直接关闭。后我们用了watch命令后台跑监控。

watch -d /bin/bash /root/sh/restart_maxsacle.sh

1
2
3
4
5
6
7
#!/bin/bash
 
netstat  -ntlp |  grep  maxscale |  grep  4006 >  /dev/null  2>&1
 
if  [ $? - eq  1 ]; then
/etc/init .d /maxscale  start
fi


解决：5.6以下可以直接在url上配置：autoReconnect=true

对于5.6以上的只能在jdbc的链接池里设置：

MySQL5.7MHA+MaxScale2.0构建高可用环境_第5张图片


第一个设置成ture,超时自动链接，对于mysql要把第二个设置成false，不启用缓存。

以上是开发需要注意的都是一些基本的配置。

 

其目的为：每次归还连接时执行select 'x'检测连接是否有效。 



——总结——

maxscale的核心就在于内个配置文件，根据不同的业务和架构，做好相应的配置，本文意在抛砖引玉，如果您有更全的配置文件，欢迎您分享给笔者。由于笔者的水平有限，编写时间也很仓促，文中难免会出现一些错误或者不准确的地方，不妥之处恳请读者批评指正。