Linux系統中多實例安裝mysql(Configuring Multiple MySQL Instances Using systemd)

一、前提條件是本地已經安裝了mysql，若是沒有安裝請參考https://dev.mysql.com/doc/refman/5.7/en/linux-installation-yum-repo.html按步驟來進行安裝。安裝完以後的話，檢測一下啓動有無問題。若是沒問題的話，接下來執行第二步驟。

二、修改/etc/my.cnf文件：內容覆蓋成以下

[mysqld@replica01]
datadir=/var/lib/mysql-replica01
socket=/var/lib/mysql-replica01/mysql.sock
port=3307
log-error=/var/log/mysqld-replica01.log

[mysqld@replica02]
datadir=/var/lib/mysql-replica02
socket=/var/lib/mysql-replica02/mysql.sock
port=3308
log-error=/var/log/mysqld-replica02.log

注意一點：此時尚未/var/lib/mysql-replica01/mysql.sock文件，也沒有/var/lib/mysql-replica02/mysql.sock文件。不要擔憂，只要你從新啓動mysql，它就會自動生成這個文件。

三、 這個時候，多實例已經配置成功了，執行如下命令讓多實例mysql啓動起來。

systemctl start mysqld@replica01
systemctl start mysqld@replica02

四、爲了能在系統啓動的時候，mysql能自動運行，須要進行以下配置

systemctl enable mysqld@replica01
systemctl enable mysqld@replica02

五、經過以下命令查看當前多實例mysql的運行狀態，可執行以下命令

systemctl status 'mysqld@replica*'

注意一下：系統命令systemctl支持通配符的形式。該命令執行的結果以下:

[root@localhost wuwenlun]# systemctl status 'mysqld@replica0*'
● mysqld@replica01.service - MySQL Server
   Loaded: loaded (/usr/lib/systemd/system/mysqld@.service; enabled; vendor preset: disabled)
   Active: failed (Result: start-limit) since Fri 2017-08-04 14:21:59 CST; 18s ago
     Docs: man:mysqld(8)
           http://dev.mysql.com/doc/refman/en/using-systemd.html
  Process: 3085 ExecStart=/usr/sbin/mysqld --defaults-group-suffix=@%I --daemonize --pid-file=/var/run/mysqld/mysqld-%i.pid $MYSQLD_OPTS (code=exited, status=1/FAILURE)
  Process: 3067 ExecStartPre=/usr/bin/mysqld_pre_systemd %I (code=exited, status=0/SUCCESS)

Aug 04 14:21:59 localhost.localdomain systemd[1]: Failed to start MySQL Server.
Aug 04 14:21:59 localhost.localdomain systemd[1]: Unit mysqld@replica01.service entered failed state.
Aug 04 14:21:59 localhost.localdomain systemd[1]: mysqld@replica01.service failed.
Aug 04 14:21:59 localhost.localdomain systemd[1]: mysqld@replica01.service holdoff time over, scheduling restart.
Aug 04 14:21:59 localhost.localdomain systemd[1]: start request repeated too quickly for mysqld@replica01.service
Aug 04 14:21:59 localhost.localdomain systemd[1]: Failed to start MySQL Server.
Aug 04 14:21:59 localhost.localdomain systemd[1]: Unit mysqld@replica01.service entered failed state.
Aug 04 14:21:59 localhost.localdomain systemd[1]: mysqld@replica01.service failed.

● mysqld@replica02.service - MySQL Server
   Loaded: loaded (/usr/lib/systemd/system/mysqld@.service; enabled; vendor preset: disabled)
   Active: failed (Result: start-limit) since Fri 2017-08-04 14:22:00 CST; 17s ago
     Docs: man:mysqld(8)
           http://dev.mysql.com/doc/refman/en/using-systemd.html
  Process: 3120 ExecStart=/usr/sbin/mysqld --defaults-group-suffix=@%I --daemonize --pid-file=/var/run/mysqld/mysqld-%i.pid $MYSQLD_OPTS (code=exited, status=1/FAILURE)
  Process: 3102 ExecStartPre=/usr/bin/mysqld_pre_systemd %I (code=exited, status=0/SUCCESS)

Aug 04 14:22:00 localhost.localdomain systemd[1]: Failed to start MySQL Server.
Aug 04 14:22:00 localhost.localdomain systemd[1]: Unit mysqld@replica02.service entered failed state.
Aug 04 14:22:00 localhost.localdomain systemd[1]: mysqld@replica02.service failed.
Aug 04 14:22:00 localhost.localdomain systemd[1]: mysqld@replica02.service holdoff time over, scheduling restart.
Aug 04 14:22:00 localhost.localdomain systemd[1]: start request repeated too quickly for mysqld@replica02.service
Aug 04 14:22:00 localhost.localdomain systemd[1]: Failed to start MySQL Server.
Aug 04 14:22:00 localhost.localdomain systemd[1]: Unit mysqld@replica02.service entered failed state.
Aug 04 14:22:00 localhost.localdomain systemd[1]: mysqld@replica02.service failed.

很遺憾，執行結果報錯，多實例myql都沒有起來。重啓系統也是如此，那麼此時先不要慌，去查看日誌先。進入/var/log/文件夾，此時你會看到 mysqld-replica01.log  mysqld-replica02.log這兩個文件，隨便選取一個文件打開，可看到以下報錯信息：

2017-08-04T06:21:58.947677Z 0 [ERROR] InnoDB: Operating system error number 13 in a file operation.
2017-08-04T06:21:58.947706Z 0 [ERROR] InnoDB: The error means mysqld does not have the access rights to the directory.
2017-08-04T06:21:58.947712Z 0 [ERROR] InnoDB: os_file_get_status() failed on './ibdata1'. Can't determine file permissions
2017-08-04T06:21:58.947719Z 0 [ERROR] InnoDB: Plugin initialization aborted with error Generic error
2017-08-04T06:21:59.550446Z 0 [ERROR] Plugin 'InnoDB' init function returned error.
2017-08-04T06:21:59.550491Z 0 [ERROR] Plugin 'InnoDB' registration as a STORAGE ENGINE failed.
2017-08-04T06:21:59.550496Z 0 [ERROR] Failed to initialize plugins.
2017-08-04T06:21:59.550498Z 0 [ERROR] Aborting

說實話看到這鬼，我也不知道是啥東西，只知道沒有權限。後來上網查了一下，只提供瞭解決方案，沒有指明具體緣由。

六、vi /etc/selinux/config修改文件爲永久關閉

# This file controls the state of SELinux on the system.
# SELINUX= can take one of these three values:
# enforcing - SELinux security policy is enforced.
# permissive - SELinux prints warnings instead of enforcing.
# disabled - No SELinux policy is loaded.
SELINUX=enforcing
# SELINUXTYPE= can take one of three two values:
# targeted - Targeted processes are protected,
# minimum - Modification of targeted policy. Only selected processes are protected. 
# mls - Multi Level Security protection.
SELINUXTYPE=targeted

SELINUX=permissive #修改此參數爲permissive 。

七、修改完以後重啓一下，執行systemctl status 'mysqld@replica0*'，得以下結果，2個mysql實例已經完美啓動了。

[root@localhost wuwenlun]# systemctl status 'mysqld@replica0*'
● mysqld@replica01.service - MySQL Server
   Loaded: loaded (/usr/lib/systemd/system/mysqld@.service; enabled; vendor preset: disabled)
   Active: active (running) since Fri 2017-08-04 14:44:25 CST; 1min 3s ago
     Docs: man:mysqld(8)
           http://dev.mysql.com/doc/refman/en/using-systemd.html
  Process: 1107 ExecStart=/usr/sbin/mysqld --defaults-group-suffix=@%I --daemonize --pid-file=/var/run/mysqld/mysqld-%i.pid $MYSQLD_OPTS (code=exited, status=0/SUCCESS)
  Process: 981 ExecStartPre=/usr/bin/mysqld_pre_systemd %I (code=exited, status=0/SUCCESS)
 Main PID: 1161 (mysqld)
   CGroup: /system.slice/system-mysqld.slice/mysqld@replica01.service
           └─1161 /usr/sbin/mysqld --defaults-group-suffix=@replica01 --daemonize --pid-file=/var/run/mysqld/mysqld-replica01.pid

Aug 04 14:44:22 localhost.localdomain systemd[1]: Starting MySQL Server...
Aug 04 14:44:25 localhost.localdomain systemd[1]: Started MySQL Server.

● mysqld@replica02.service - MySQL Server
   Loaded: loaded (/usr/lib/systemd/system/mysqld@.service; enabled; vendor preset: disabled)
   Active: active (running) since Fri 2017-08-04 14:44:25 CST; 1min 3s ago
     Docs: man:mysqld(8)
           http://dev.mysql.com/doc/refman/en/using-systemd.html
  Process: 1095 ExecStart=/usr/sbin/mysqld --defaults-group-suffix=@%I --daemonize --pid-file=/var/run/mysqld/mysqld-%i.pid $MYSQLD_OPTS (code=exited, status=0/SUCCESS)
  Process: 973 ExecStartPre=/usr/bin/mysqld_pre_systemd %I (code=exited, status=0/SUCCESS)
 Main PID: 1159 (mysqld)
   CGroup: /system.slice/system-mysqld.slice/mysqld@replica02.service
           └─1159 /usr/sbin/mysqld --defaults-group-suffix=@replica02 --daemonize --pid-file=/var/run/mysqld/mysqld-replica02.pid

Aug 04 14:44:22 localhost.localdomain systemd[1]: Starting MySQL Server...
Aug 04 14:44:25 localhost.localdomain systemd[1]: Started MySQL Server.

八、可經過以下命令進入指定的mysql

mysql -uroot -p -S /var/lib/mysql-replica01/mysql.sock
 mysql -uroot -p -S /var/lib/mysql-replica02/mysql.sock

九、這個時候，你沒有密碼進入不了。那麼執行以下命令可獲得mysql安裝的時候隨機生成的一個初始密碼

sudo grep 'temporary password' /var/log/mysqld-replica01.log
sudo grep 'temporary password' /var/log/mysqld-replica02.log