mysql HA
MySQL5.7+Keepalived双主HA架构
https://blog.csdn.net/cuiruidu3106/article/details/100438947
赋予执行权限
HA部署步骤
1、mysql_check.sh脚本
发现mysql挂了 杀死本机的keepalived
主备均部署
cat > /etc/keepalived/mysql_check.sh << "EOF" #!/bin/bash counter=$(netstat -na|grep "LISTEN"|grep "3306"|wc -l) if [ "${counter}" -eq 0 ]; then echo "$(date +'%F %T') mysql down, keepalived will stop" >> /var/log/keepalived.log systemctl stop keepalived fi EOF
chmod +x /etc/keepalived/mysql_check.sh
2、Master配置文件编写
只主部署
cat >/etc/keepalived/keepalived.conf << EOF global_defs { router_id mysql_01 # 设置master的id,在一个网络应该是唯一的 } vrrp_script chk_script { script "/etc/keepalived/mysql_check.sh" #最后手动执行下此脚本,以确保此脚本能够正常执行 interval 2 #(检测脚本执行的间隔,单位是秒) weight 20 } vrrp_instance VI_1 { state BACKUP nopreempt priority 100 # 优先级,数值越大,获取处理请求的优先级越高 advert_int 1 # 检查间隔,默认为1s(vrrp组播周期秒数) virtual_router_id 77 # 虚拟路由编号,主从要一致 interface eth0 # 当前进行vrrp通讯的网络接口卡(当前centos的网卡) authentication { auth_type PASS auth_pass yunzhihui123 } track_script { chk_script #(调用检测脚本) } virtual_ipaddress { 172.16.1.254 # 定义虚拟ip(VIP),可多设,每行一个 } } EOF
3、Backup配置文件编写
只备部署
cat >/etc/keepalived/keepalived.conf << EOF global_defs { router_id mysql_02 # 设置backup的id,在一个网络应该是唯一的 } vrrp_script chk_script { script "/etc/keepalived/mysql_check.sh" #最后手动执行下此脚本,以确保此脚本能够正常执行 interval 2 #(检测脚本执行的间隔,单位是秒) weight 20 } vrrp_instance VI_1 { state BACKUP nopreempt priority 90 # 优先级,数值越大,获取处理请求的优先级越高 advert_int 1 # 检查间隔,默认为1s(vrrp组播周期秒数) virtual_router_id 77 # 虚拟路由编号,主从要一致 interface eth0 # 当前进行vrrp通讯的网络接口卡(当前centos的网卡) authentication { auth_type PASS auth_pass yunzhihui123 } track_script { chk_script #(调用检测脚本) } virtual_ipaddress { 172.16.1.254 # 定义虚拟ip(VIP),可多设,每行一个 } } EOF
4、配置文件生效
systemctl stop keepalived
systemctl start keepalived
HA验证
1、查看VIP在哪台机器
ip a
VIP 目前在主,备无VIP
如果两台都有VIP,请检查防火墙,确保只有一台有VIP
3、杀死主上的Mysql
推荐systemctl stop mysql
4、验证VIP漂移
VIP 现在跑到了备,符合VIP现状
6、HA验证完毕后恢复原状
主 备 启动Mysql (先) 和 Keepalived (后)
其他
恢复
Mysql启动后(先),可以重新启动Keepalived(后),进行恢复
如果发现VIP正常,但打不开网页,可能网关arp没有刷新,在VIP所在机器上执行以下命令一两次
172.16.1.254 为VIP
172.16.1.1 为网关
查看网关
netstat -rn
(以0.0.0.0开始的行的gateway是默认网关)
/sbin/arping -I eth0 -c 5 -s 172.16.1.254 172.16.1.1 &>/dev/null
HA部署步骤 1、mysql_check.sh脚本 发现mysql挂了 杀死本机的keepalived 主备均部署 cat > /etc/keepalived/check_mysql.sh << "EOF" #!/bin/bash MYSQL=/data/app/mysql/bin/mysql MYSQL_HOST=localhost MYSQL_USER=root MYSQL_PASSWORD=yunzhihui123 CHECK_TIME=3 #mysql is working MYSQL_OK is 1 , mysql down MYSQL_OK is 0 MYSQL_OK=1 function check_mysql_helth (){ $MYSQL -h $MYSQL_HOST -u $MYSQL_USER -p${MYSQL_PASSWORD} -e "select 1 from dual;" >/dev/null 2>&1 if [ $? = 0 ] ;then MYSQL_OK=1 else MYSQL_OK=0 fi return $MYSQL_OK } while [ $CHECK_TIME -ne 0 ] do let "CHECK_TIME -= 1" check_mysql_helth if [ $MYSQL_OK = 1 ] ; then CHECK_TIME=0 exit 0 fi if [ $MYSQL_OK -eq 0 ] && [ $CHECK_TIME -eq 0 ] then echo "$(date +'%F %T') mysql down, keepalived will stop" >> /var/log/keepalived.log pkill keepalived exit 1 fi sleep 1 done EOF 赋予执行权限 chmod u+x /etc/keepalived/check_mysql.sh 2、Master配置文件编写 只主部署 cat >/etc/keepalived/keepalived.conf << EOF global_defs { router_id mysql_01 # 设置master的id,在一个网络应该是唯一的 } vrrp_script chk_script { script "/etc/keepalived/check_mysql.sh" #最后手动执行下此脚本,以确保此脚本能够正常执行 interval 5 #(检测脚本执行的间隔,单位是秒) weight 20 } vrrp_instance VI_1 { state BACKUP nopreempt priority 100 # 优先级,数值越大,获取处理请求的优先级越高 advert_int 1 # 检查间隔,默认为1s(vrrp组播周期秒数) virtual_router_id 77 # 虚拟路由编号,主从要一致 interface eth0 # 当前进行vrrp通讯的网络接口卡(当前centos的网卡) authentication { auth_type PASS auth_pass mysqlyunzhihui123 } track_script { chk_script #(调用检测脚本) } virtual_ipaddress { 172.16.1.254 # 定义虚拟ip(VIP),可多设,每行一个 } notify_master /etc/keepalived/notify_master_mysql.sh } EOF 3、Backup配置文件编写 只备部署 cat >/etc/keepalived/keepalived.conf << EOF global_defs { router_id mysql_02 # 设置backup的id,在一个网络应该是唯一的 } vrrp_script chk_script { script "/etc/keepalived/check_mysql.sh" #最后手动执行下此脚本,以确保此脚本能够正常执行 interval 5 #(检测脚本执行的间隔,单位是秒) weight 20 } vrrp_instance VI_1 { state BACKUP nopreempt priority 90 # 优先级,数值越大,获取处理请求的优先级越高 advert_int 1 # 检查间隔,默认为1s(vrrp组播周期秒数) virtual_router_id 77 # 虚拟路由编号,主从要一致 interface eth0 # 当前进行vrrp通讯的网络接口卡(当前centos的网卡) authentication { auth_type PASS auth_pass mysqlyunzhihui123 } track_script { chk_script #(调用检测脚本) } virtual_ipaddress { 172.16.1.254 # 定义虚拟ip(VIP),可多设,每行一个 } notify_master /etc/keepalived/notify_master_mysql.sh } EOF 4、配置文件生效 systemctl stop keepalived systemctl start keepalived HA验证 1、查看VIP在哪台机器 ip a VIP 目前在主,备无VIP 如果两台都有VIP,请检查防火墙,确保只有一台有VIP 3、杀死主上的Mysql 推荐systemctl stop mysql 4、验证VIP漂移 VIP 现在跑到了备,符合VIP现状 6、HA验证完毕后恢复原状 主 备 启动Mysql (先) 和 Keepalived (后) 其他 恢复 Mysql启动后(先),可以重新启动Keepalived(后),进行恢复 如果发现VIP正常,但打不开网页,可能网关arp没有刷新,在VIP所在机器上执行以下命令一两次 172.16.1.254 为VIP 172.16.1.1 为网关 查看网关 netstat -rn (以0.0.0.0开始的行的gateway是默认网关) /sbin/arping -I eth0 -c 5 -s 172.16.1.254 172.16.1.1 &>/dev/null 关于SUSE 12防火墙 关于SUSE 12防火墙