官網(wǎng): https://github.com/yoshinorim/mha4mysql-manager/wiki
前提:
?? ?1) 所有節(jié)點之間, 相互之間免密認證
?? ?2) 所有節(jié)點yum源配置正常
?? ?3) gw服務器有相關的安裝包
?? ?4) 系統(tǒng)時間要正確
?? ?5) /etc/hosts要統(tǒng)一(本實驗通過dns解析已經(jīng)實現(xiàn))
環(huán)境拓撲:
node01 ?10.15.200.101 ?Primary_Master ? ServerID=101 ?主(讀寫)
node02 ?10.15.200.102 ?Candidate_Master ServerID=102 ?從(讀)
node03 ?10.15.200.103 ?slave01 ? ? ? ? ?ServerID=103 ?從(讀)
node05 ?10.15.200.105 ?mha ? ? ? ? ? ? ?ServerID=105 ?高可用監(jiān)控
node01為主服務器 node02為從服務器 node03為從服務器
當主服務器:node01故障時 node02由從提升為主 node03將主服務器重新指向node02(原為node01)
# 1. 所有節(jié)點(node01 node02 node03 node05) 清空之前實驗的相關數(shù)據(jù) ?(刪除之前實驗環(huán)境數(shù)據(jù) 清空yum舊緩存 更新yum新緩存
# 或是 直接還原快照
yum remove Percona-XtraDB-Cluster* httpd* php* Percona-Server* mariadb* -y
yum remove mysql-community-server -y
rm -fr /var/log/mysqld.log /var/lib/mysql/*
yum clean all && yum makecache
# 2. 在node01 node02 node03 安裝mysql-community-5.7
yum -y install mysql-community-server
?
# 數(shù)據(jù)庫配置文件已經(jīng)放在gw (如果沒有 請自行上傳)
node01: wget http://10.15.200.8/files/mha/node01.my.cnf -O /etc/my.cnf
node02: wget http://10.15.200.8/files/mha/node02.my.cnf -O /etc/my.cnf
node03: wget http://10.15.200.8/files/mha/node03.my.cnf -O /etc/my.cnf
或者是這樣的一條命令：
wget http://10.15.200.8/files/mha/`hostname -s`.my.cnf -O /etc/my.cnf
# 3. 在所有節(jié)點 安裝MHA (node01 node02 node03 node05)
yum install mha4mysql-node-0.58 -y
?
主節(jié)點: node01 安裝數(shù)據(jù)庫 啟動服務 ?配置用戶名和密碼
[root@node01 ~]# rm -fr /var/lib/mysql/* && mysqld --initialize && chown -R mysql:mysql /var/lib/mysql
[root@node01 ~]# systemctl start mysqld
[root@node01 ~]# grep 'temporary password' /var/log/mysqld.log | awk '{print $11}'
Fc-8L0tu?rMh
# 修改root密碼 創(chuàng)建用戶復制用戶repl及mha管理用戶mha
[root@node01 ~]# mysql -uroot -p'Fc-8L0tu?rMh'
ALTER USER 'root'@'localhost' IDENTIFIED BY '!@#qweASD69';
grant replication slave on *.* ?to repl@'10.15.200.%' identified by '!@#qweASD69';
grant all on *.* to mha@'10.15.200.%' identified by '!@#qweASD69';
flush privileges;
因為要做主從實驗，所以各個服務器的server-id不能相同，我們需要在這里進行查看，確保沒有問題
?
?
node02: 同理類似
[root@node02 ~]# rm -fr /var/lib/mysql/* && mysqld --initialize --user=mysql
[root@node02 ~]# systemctl start mysqld
[root@node02 ~]# grep 'temporary password' /var/log/mysqld.log | awk '{print $11}'
G:aFM:4tkn4y
[root@node02 ~]# mysql -uroot -p'G:aFM:4tkn4y'
ALTER USER 'root'@'localhost' IDENTIFIED BY '!@#qweASD69';
grant replication slave on *.* ?to repl@'10.15.200.%' identified by '!@#qweASD69';
grant all on *.* to mha@'10.15.200.%' identified by '!@#qweASD69';
flush privileges;
node02上配置從服務器:
CHANGE MASTER TO
MASTER_HOST='10.15.200.101',
MASTER_PORT=3306,
MASTER_AUTO_POSITION=1,
MASTER_USER='repl',
MASTER_PASSWORD='!@#qweASD69';
mysql> show slave status\G;
? ? ? ? ? ? ?Slave_IO_Running: No
? ? ? ? ? ? Slave_SQL_Running: No
mysql> start slave;
mysql> show slave status\G;
*************************** 1. row ***************************
? ? ? ? ? ? ? ?Slave_IO_State: Waiting for master to send event
? ? ? ? ? ? ? ? ? Master_Host: 10.15.200.101
? ? ? ? ? ? ? ? ? Master_User: repl
? ? ? ? ? ? ? ? ? Master_Port: 3306
? ? ? ? ? ? 重點是這里，得有
node03: 同理類似
[root@node03 ~]# rm -fr /var/lib/mysql/* && mysqld --initialize --user=mysql
[root@node03 ~]# systemctl start mysqld
[root@node03 ~]# grep 'temporary password' /var/log/mysqld.log | awk '{print $11}'
>uLMJdwI/1Tl
[root@node03 ~]# mysql -uroot -p'>uLMJdwI/1Tl'
ALTER USER 'root'@'localhost' IDENTIFIED BY '!@#qweASD69';
grant replication slave on *.* ?to repl@'10.15.200.%' identified by '!@#qweASD69';
grant all on *.* to mha@'10.15.200.%' identified by '!@#qweASD69';
flush privileges;
node03上配置從服務器:
CHANGE MASTER TO
MASTER_HOST='10.15.200.101',
MASTER_PORT=3306,
MASTER_AUTO_POSITION=1,
MASTER_USER='repl',
MASTER_PASSWORD='!@#qweASD69';
mysql> show slave status\G;
? ? ? ? ? ? ?Slave_IO_Running: No
? ? ? ? ? ? Slave_SQL_Running: No
mysql> start slave;
mysql> show slave status\G;
*************************** 1. row ***************************
? ? ? ? ? ? ? ?Slave_IO_State: Waiting for master to send event
? ? ? ? ? ? ? ? ? Master_Host: 10.15.200.101
? ? ? ? ? ? ? ? ? Master_User: repl
? ? ? ? ? ? ? ? ? Master_Port: 3306
? ? ? ? ? ? ? ? Connect_Retry: 60
# ping虛擬ip 無法ping通
~$ ping 10.15.200.118
PING 10.15.200.118 (10.15.200.118): 56 data bytes
Request timeout for icmp_seq 0
node01: 先在主庫master上綁定VIP（只需手工綁定一次,后續(xù)腳本會自動切換）
[root@node01 ~]# ifconfig ens33:1 10.15.200.118/24 ? ? ?# 注意如果是ens32做對應的修改
[root@node01 ~]# ip addr
2: ens33: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP group default qlen 1000
? ? inet 10.15.200.118/24 brd 10.15.200.255 scope global secondary ens33:1
? ? ? ?valid_lft forever preferred_lft forever
? ? ? ?
~$ ping 10.15.200.118 -c2
PING 10.15.200.118 (10.15.200.118): 56 data bytes
64 bytes from 10.15.200.118: icmp_seq=0 ttl=64 time=1.431 ms
# node05安裝管理節(jié)點
yum install mha4mysql-node-0.58 mha4mysql-manager-0.58 -y
mkdir -p /etc/mha/scripts
wget http://10.15.200.8/files/mha/master_ip_failover -O /etc/mha/scripts/master_ip_failover
wget http://10.15.200.8/files/mha/master_ip_online_change -O /etc/mha/scripts/master_ip_online_change
wget http://10.15.200.8/files/mha/send_report -O /etc/mha/scripts/send_report
chmod 744 /etc/mha/scripts/*
wget http://10.15.200.8/files/mha/app.cnf -O /etc/mha/app.cnf
wget http://10.15.200.8/files/mha/masterha_default.cnf -O /etc/masterha_default.cnf
node05:
/etc/mha/scripts/master_ip_failover ? ?? ??? ?# 修改網(wǎng)址的名字 ?ens32 ens34 ens33
/etc/mha/scripts/master_ip_online_change ? ?# 修改網(wǎng)址的名字 ?ens32 ens34 ens33
node05賦予腳本執(zhí)行權限:
[root@node05 ~]# chmod 744 /etc/mha/scripts/*
# 用 masterha_check_ssh 命令檢查 ssh 互信是否成功
[root@node05 ~]# masterha_check_ssh --conf=/etc/mha/app.cnf
Tue Mar ?2 23:08:01 2021 - [info] All SSH connection tests passed successfully.
# 使用 masterha_check_repl 命令檢查 mysql 主從是否正常
[root@node05 ~]# masterha_check_repl --conf=/etc/mha/app.cnf
MySQL Replication Health is OK.
啟動MHA: 如下命令 會卡住不動
[root@node05 ~]# masterha_manager --conf=/etc/mha/app.cnf --remove_dead_master_conf --ignore_last_failover < /dev/null > /var/log/mha/app/manager.log 2>&1
通過日志檢查MHA是否啟動成功:
[root@node05 ~]# tailf /var/log/mha/app/manager.log
Checking the Status of the script.. OK
Tue Mar ?2 23:15:30 2021 - [info] ?OK.
Tue Mar ?2 23:15:30 2021 - [warning] shutdown_script is not defined.
Tue Mar ?2 23:15:30 2021 - [info] Set master ping interval 1 seconds.
Tue Mar ?2 23:15:30 2021 - [info] Set secondary check script: /usr/bin/masterha_secondary_check -s node01 -s node02 -s node03
Tue Mar ?2 23:15:30 2021 - [info] Starting ping health check on node01(10.15.200.101:3306)..
Tue Mar ?2 23:15:30 2021 - [info] Ping(SELECT) succeeded, waiting until MySQL doesn't respond..
# 最后一行出現(xiàn)如下字樣表明啟動成功
Tue Mar ?2 23:15:30 2021 - [info] Ping(SELECT) succeeded, waiting until MySQL doesn't respond..
Mon Sep 20 23:43:27 2021 - [info] Got exit code 1 (Not master dead).
檢查MHA集群狀態(tài):
[root@node05 ~]# masterha_check_status --conf=/etc/mha/app.cnf
app (pid:3412) is running(0:PING_OK), master:node01
將node01的mysql服務停止:
[root@node01 ~]# systemctl stop mysqld
node05動態(tài)的查看日志:
[root@node05 scripts]# tailf /var/log/mha/app/manager.log
IN SCRIPT TEST====/sbin/ifconfig ens33:1 down==/sbin/ifconfig ens33:1 10.15.200.118/24===
Enabling the VIP - 10.15.200.118/24 on the new master - node02
Warning: Permanently added 'node02,10.15.200.102' (ECDSA) to the list of known hosts.
Tue Mar ?2 23:21:56 2021 - [info] ?OK.
Tue Mar ?2 23:21:56 2021 - [info] ** Finished master recovery successfully.
Tue Mar ?2 23:21:56 2021 - [info] * Phase 3: Master Recovery Phase completed.
Tue Mar ?2 23:21:56 2021 - [info]
Tue Mar ?2 23:21:56 2021 - [info] * Phase 4: Slaves Recovery Phase..
Tue Mar ?2 23:21:56 2021 - [info]
Tue Mar ?2 23:21:56 2021 - [info]
Tue Mar ?2 23:21:56 2021 - [info] * Phase 4.1: Starting Slaves in parallel..
Tue Mar ?2 23:21:56 2021 - [info]
Tue Mar ?2 23:21:56 2021 - [info] -- Slave recovery on host node03(10.15.200.103:3306) started, pid: 3884. Check tmp log /var/log/mha/app/node03_3306_20210302232154.log if it takes time..
Tue Mar ?2 23:21:58 2021 - [info]
Tue Mar ?2 23:21:58 2021 - [info] Log messages from node03 ...
Tue Mar ?2 23:21:58 2021 - [info]
Tue Mar ?2 23:21:56 2021 - [info] ?Resetting slave node03(10.15.200.103:3306) and starting replication from the new master node02(10.15.200.102:3306)..
Tue Mar ?2 23:21:56 2021 - [info] ?Executed CHANGE MASTER.
Tue Mar ?2 23:21:57 2021 - [info] ?Slave started.
Tue Mar ?2 23:21:57 2021 - [info] ?gtid_wait(bfb0229c-7b5b-11eb-9393-000c29edf7ca:1-6,
eec2d10b-7b5e-11eb-afcc-000c2930884c:1-3) completed on node03(10.15.200.103:3306). Executed 0 events.
Tue Mar ?2 23:21:58 2021 - [info] End of log messages from node03.
Tue Mar ?2 23:21:58 2021 - [info] -- Slave on host node03(10.15.200.103:3306) started.
Tue Mar ?2 23:21:58 2021 - [info] All new slave servers recovered successfully.
Tue Mar ?2 23:21:58 2021 - [info]
Tue Mar ?2 23:21:58 2021 - [info] * Phase 5: New master cleanup phase..
Tue Mar ?2 23:21:58 2021 - [info]
Tue Mar ?2 23:21:58 2021 - [info] Resetting slave info on the new master..
Tue Mar ?2 23:21:58 2021 - [info] ?node02: Resetting slave info succeeded.
Tue Mar ?2 23:21:58 2021 - [info] Master failover to node02(10.15.200.102:3306) completed successfully.
Tue Mar ?2 23:21:58 2021 - [info] Deleted server1 entry from /etc/mha/app.cnf .
Tue Mar ?2 23:21:58 2021 - [info]
----- Failover Report -----
app: MySQL Master failover node01(10.15.200.101:3306) to node02(10.15.200.102:3306) succeeded
Master node01(10.15.200.101:3306) is down!
Check MHA Manager logs at node05.example.cn:/var/log/mha/app/manager.log for details.
Started automated(non-interactive) failover.
Invalidated master IP address on node01(10.15.200.101:3306)
Selected node02(10.15.200.102:3306) as a new master.
node02(10.15.200.102:3306): OK: Applying all logs succeeded.
node02(10.15.200.102:3306): OK: Activated master IP address.
node03(10.15.200.103:3306): OK: Slave started, replicating from node02(10.15.200.102:3306)
node02(10.15.200.102:3306): Resetting slave info succeeded.
Master failover to node02(10.15.200.102:3306) completed successfully.
# 虛擬IP已經(jīng)漂移到node02 ?node02由從的角色 提升為主的角色
[root@node02 mysql]# ip addr
? ? inet 10.15.200.118/24 brd 10.15.200.255 scope global secondary ens33:1
? ? ? ?valid_lft forever preferred_lft forever
# 此時發(fā)現(xiàn) node03的主服務器由node01變?yōu)閚ode02
[root@node03 ~]# mysql -uroot -p'!@#qweASD' -e 'show slave status\G;'
mysql: [Warning] Using a password on the command line interface can be insecure.
*************************** 1. row ***************************
? ? ? ? ? ? ? ?Slave_IO_State: Waiting for master to send event
? ? ? ? ? ? ? ? ? Master_Host: 10.15.200.102
? ? ? ? ? ? ? ? ??
此時再次查看集群的狀態(tài) 為失效的狀態(tài):
[root@node05 ~]# masterha_check_status --conf=/etc/mha/app.cnf
app is stopped(2:NOT_RUNNING).
[root@node05 ~]# cat /etc/mha/app.cnf ? ?# 此文件中的[server1]段內(nèi)容已經(jīng)被刪除
注意事項:
?? ?1) master端同樣要開啟兩個重要的選項,server-id和log-bin
?? ??? ?并且server-id在全局架構中并且唯一,不能被其它主機使用
?? ??? ?slave端要開啟relay-log
https://github.com/yoshinorim/mha4mysql-manager/wiki文章來源：http://www.zghlxwxcb.cn/news/detail-424980.html
ERROR 1819 (HY000): Your password does not satisfy the current policy requirements
ERROR 1820 (HY000): You must reset your password using ALTER USER statement before executing this statement
?文章來源地址http://www.zghlxwxcb.cn/news/detail-424980.html
到了這里，關于云計算基礎架構（一） 數(shù)據(jù)庫的文章就介紹完了。如果您還想了解更多內(nèi)容，請在右上角搜索TOY模板網(wǎng)以前的文章或繼續(xù)瀏覽下面的相關文章，希望大家以后多多支持TOY模板網(wǎng)！
![[架構之路-236]：目標系統(tǒng) - 縱向分層 - 數(shù)據(jù)庫 - 數(shù)據(jù)庫系統(tǒng)基礎與概述：三階段模型（概念模型、邏輯模型、物理模型）、三級模式結構（外模式、模式、內(nèi)模式）](https://imgs.yssmx.com/Uploads/2024/02/782939-1.png)
