环境及软件版本：
CentOS6.5x86_64
MysqL5.6.34编译安装版
MHA版本：mha4MysqL-manager-0.56-0.el6.noarch.rpm mha4MysqL-node-0.56-0.el6.noarch.rpm
节点角色：
node93：10.1.20.93 默认主库
node94：10.1.20.94 从库1，原先主库宕机后可提升为主库【mha管理节点也部署在这台机器上】
node95：10.1.20.95 从库2，不允许提升为主库
准备的VIP是 10.1.20.100/24
step1、配置主从关系
node93的/etc/my.cnf配置文件部分关键地方：
[MysqLd]
port = 3306
socket = /tmp/MysqL.sock
datadir = /bdata/data/nowdb2
innodb_file_per_table=ON
character-set-server = utf8
default_storage_engine = InnoDB
skip-innodb_adaptive_hash_index
master_info_repository = TABLE
relay_log_info_repository = TABLE
relay_log_recovery = 1 #crash safe
log-bin=MysqL-bin
binlog_format=row
sync_binlog =1 #确保事务提交的时候BINLOG落盘
log-slave-updates
log_bin_trust_function_creators =1
binlog_rows_query_log_events=ON #记录执行的语句到BINLOG query event
server-id=1020093
relay_log_purge=0
read_only=1
node94的/etc/my.cnf配置文件部分关键地方：
server-id=1020094
node95的/etc/my.cnf配置文件部分关键地方：
log_bin_trust_function_creators = 1
server-id=1020095
在node93上创建复制权限的账号，GRANT REPLICATION SLAVE,REPLICATION CLIENT ON *.* TO 'rpl'@'10.1.%.%' IDENTIFIED BY 'rpl';
然后配置1主2从，（具体步骤略过）。
注意：我们要确保能成为主库的节点(node93、node94)都存在主从同步账号，如果node94上不存在rpl账号，就到节点去手工添加即可。
主从关系建立好后，我们在master上创建个mha管理账号，后期会用到：
grant all on *.* to 'mhauser'@'10.1.%.%' identified by 'Abcd@1234';
(管理账号要在node93、node94、node95所有节点都存在)
step2、安装MHA
因为MHA依赖于SSH，因此需要在3台主机之间建立SSH免秘钥登陆。步骤略过。
3个节点都安装perl包：
yum install perl perl-DBD-MysqL perl-CPAN perl-devel perl-Time-HiRes
node93-node95上都安装node包：
rpm -ivh mha4MysqL-node-0.56-0.el6.noarch.rpm
node94上安装Manager包(当然，我们在3个节点都安装上Manager包也没问题)：
rpm -ivh mha4MysqL-manager-0.56-0.el6.noarch.rpm
在node94初始化MHA
mkdir /etc/masterha/
vim /etc/masterha/app1.cnf 内容如下：
[server default]
user=mhauser
password=Abcd@1234
manager_workdir=/data/masterha/app1
manager_log=/data/masterha/app1/manager.log
remote_workdir=/data/masterha/app1
ssh_user=root
repl_user=rpl
repl_password=rpl
ping_interval=1
master_binlog_dir=/bdata/data/nowdb2/ # 这个路径和你的MysqL的binlog存放路径要一致
master_ip_failover_script=/etc/masterha/master_ip_failover
report_script=/etc/masterha/send_report
#通过第三方机器确认目标主库是否存活,不是必须的,就算没有也是能用
#secondary_check_script=masterha_secondary_check -s remote_host1 -s remote_host2
#故障发生后关闭主机的脚本,但是你要设置为空
# shutdown_script=""
#手动在线切换VIP脚本,就算没有也是能用,
#如果你有keepalived这种来做切换VIP就可以直接不用了
master_ip_online_change_script==/etc/masterha/master_ip_online_change
[server1]
hostname=10.1.20.93
candidate_master=1
[server2]
hostname=10.1.20.94
[server3]
hostname=10.1.20.95
no_master=1 # 禁止
在node94上添加脚本/etc/masterha/master_ip_failover (里面填写上相关的VIP的信息)
#!/usr/bin/env perl
use strict;
use warnings FATAL => 'all';
use Getopt::Long;
my (
$command,$ssh_user,$orig_master_host,$orig_master_ip,
$orig_master_port,$new_master_host,$new_master_ip,$new_master_port,238);">$orig_master_ssh_port,$new_master_ssh_port
);
my $vip ='10.1.20.100'; # Virtual IP
my $devic='eth0';
my $key = "0";
my$net_mask='255.255.255.0';
my $ssh_start_vip ="/sbin/ifconfig $devic:$key $vip netmask $net_mask";
my $ssh_stop_vip ="/sbin/ifconfig $devic:$key down";
my$MysqL_conf="/etc/my.cnf";
my $open_readonly="/bin/sed-i 's/.*read_only.*/read_only=1/g' $MysqL_conf ";
my $close_readonly="/bin/sed-i 's/.*read_only.*/read_only=0/g' $MysqL_conf ";
my $open_relaylog_purge="/bin/sed-i 's/.*relay_log_purge.*/relay_log_purge=0/g' $MysqL_conf ";
my$close_relaylog_purge="/bin/sed -i's/.*relay_log_purge.*/relay_log_purge=0/g' $MysqL_conf ";
GetOptions(
'command=s' =>\$command,238);">'ssh_user=s' =>\$ssh_user,238);">'orig_master_host=s' => \$orig_master_host,238);">'orig_master_ip=s' =>\$orig_master_ip,238);">'orig_master_port=i' => \$orig_master_port,238);">'orig_master_ssh_port=i' => \$orig_master_ssh_port,238);">'new_master_host=s' => \$new_master_host,238);">'new_master_ip=s' =>\$new_master_ip,238);">'new_master_port=i' =>\$new_master_port,238);">'new_master_ssh_port=i' => \$new_master_ssh_port,238);">exit &main();
sub main {
print "\n\nIN SCRIPTTEST====$ssh_stop_vip==$ssh_start_vip===\n\n";
if ( $command eq "stop" || $command eq "stopssh" ) {
# $orig_master_host,$orig_master_port are passed.
# If you manage master ip address atglobal catalog database,238);"> # invalidate orig_master_ip here.
my $exit_code = 1;
eval {
print "Disabling the VIP onold master: $orig_master_host \n";
&stop_vip();
$exit_code = 0;
};
if ($@) {
warn "Got Error: $@\n";
exit $exit_code;
}
}
elsif ( $command eq "start" ) {
# all arguments are passed.
# activate new_master_ip here.
# You can also grant write access(create user,set read_only=0,etc) here.
my $exit_code = 10;
print "Enabling the VIP - $vipon the new master - $new_master_host \n";
&start_vip();
warn $@;
elsif ( $command eq "status" ) {
print "Checking the Status of thescript.. OK \n";
# `ssh $ssh_user\@cluster1 \"$ssh_start_vip \"`;
exit 0;
else {
&usage();
exit 1;
# A simple system call that enablethe VIP on the new master
sub start_vip() {
`ssh $ssh_user\@$new_master_host \" $ssh_start_vip \"`;
print "Disable read_only and relay_log_purge in my.cnf - on the new master - $new_master_host\n";
`ssh $ssh_user\@$new_master_host \" $close_readonly \"`;
`ssh $ssh_user\@$new_master_host \" $close_relaylog_purge \"`;
# A simple system call thatdisable the VIP on the old_master
sub stop_vip() {
`ssh $ssh_user\@$orig_master_host \" $ssh_stop_vip \"`;
print "Enable read_only and relay_log_purge in my.cnf - on the orig master - $orig_master_host\n";
`ssh $ssh_user\@$orig_master_host \" $open_readonly \"`;
`ssh $ssh_user\@$orig_master_host \" $open_relaylog_purge \"`;
sub usage {
"Usage: master_ip_failover --command=start|stop|stopssh|status--orig_master_host=host --orig_master_ip=ip --orig_master_port=port--new_master_host=host --new_master_ip=ip --new_master_port=port--orig_master_ssh_port=ssh_port --new_master_ssh_port = ssh_port\n";
}
在node94上添加脚本/etc/masterha/send_report(里面填写上相关的smtp账号的信息)：
#!/usr/bin/perl
#Copyright (C) 2011 DeNA Co.,Ltd.
#
#This program is free software; you can redistribute it and/or modify
#it under the terms of the GNU General Public License as published by
#the Free Software Foundation; either version 2 of the License,or
#(at your option) any later version.
#This program is distributed in the hope that it will be useful,238);">#but WITHOUT ANY WARRANTY; without even the implied warranty of
#MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
#GNU General Public License for more details.
#You should have received a copy of the GNU General Public License
#along with this program; if not,write to the Free Software
#Foundation,Inc.,238);">#51 Franklin Street,Fifth Floor,Boston,MA 02110-1301USA
## Note: This is a sample scriptand is not complete. Modify the script based on your environment.
use Mail::Sender;
#new_master_host and new_slave_hostsare set only when recovering master succeeded
my ( $dead_master_host,$new_slave_hosts,$subject,$body,$conf);
my$smtp='smtp.exmail.qq.com';
my$mail_from='zabbix@xxxx.com';
my$mail_user='zabbix@xxxx.com';
my $mail_pass='xxxxxxx';
my$mail_to=['lixxxx@xxxx.com'];
'orig_master_host=s' => \$dead_master_host,238);">'new_master_host=s' =>\$new_master_host,238);">'new_slave_hosts=s' =>\$new_slave_hosts,238);">'subject=s' =>\$subject,238);">'body=s' => \$body,238);">'conf=s' => \$conf,238);">mailToContacts($smtp,$mail_from,$mail_user,$mail_pass,$mail_to,$body);
check_if_sendmail_ok('/tmp/monitormail.log');
sub mailToContacts {
my ( $smtp,$user,$passwd,$msg ) = @_;
open my $DEBUG,"> /tmp/monitormail.log"
or die "Can't open the debug file:$!\n";
my $sender = new Mail::Sender {
ctype => 'text/plain; charset=utf-8',238);"> encoding => 'utf-8',238);"> smtp => $smtp,238);"> from => $mail_from,238);"> auth => 'LOGIN',238);"> TLS_allowed => '0',238);"> authid => $user,238);"> authpwd => $passwd,238);"> to => $mail_to,238);"> subject => $subject,238);"> debug => $DEBUG
};
$sender->MailMsg(
{msg => $msg,238);">) or print $Mail::Sender::Error;
return 1;
sub check_if_sendmail_ok{
#>>250 2.0.0 Ok: queued as 3532C6DA009D
#<<QUIT
#>>221 2.0.0 Bye
my$logf = shift;
openRLOG,$logf or die "cannot open file $logf.\n";
my@log = <RLOG>;
closeRLOG;
my$val = 0;
if($log[$#log]=~ m/\>\>\s221\s.*\sBye/){
print"Meet Bye.\t";
$val++;
if($log[$#log-1]=~ m/\<\<\sQUIT/){
print"Meet QUIT.\t";
if($log[$#log-2]=~ m/\>\>\s250\s.*\sOk: queued/){
print"Meet queued.\t";
print"\n";
if($val== 3){
print"send mail success.\n";
else{
print"send mail Failed.check DNS/SMTP config\n";
return$val;
# Do whatever you want here
exit 0;
在node94上添加脚本/etc/masterha/master_ip_online_change(里面填写上相关的VIP的信息)：
use MHA::DBHelper;
use MHA::NodeUtil;
use Time::HiRes qw( sleepgettimeofday tv_interval );
use Data::Dumper;
my $_tstart;
my $_running_interval = 0.1;
$orig_master_ip,$orig_master_port,$orig_master_user,238);">$orig_master_password,$orig_master_ssh_user,238);">$new_master_ip,$new_master_user,238);">$new_master_password,$new_master_ssh_user
my $vip ='10.1.20.100/24';
my $key = '0';
my $ssh_start_vip ="/sbin/ifconfig eth0:$key $vip";
my $ssh_stop_vip ="/sbin/ifconfig eth0:$key down";
my $orig_master_ssh_port = 22;
my $new_master_ssh_port = 22;
'orig_master_is_new_slave' => \$orig_master_is_new_slave,238);">'orig_master_host=s' =>\$orig_master_host,238);">'orig_master_port=i' =>\$orig_master_port,238);">'orig_master_user=s' =>\$orig_master_user,238);">'orig_master_password=s' =>\$orig_master_password,238);">'orig_master_ssh_user=s' =>\$orig_master_ssh_user,238);">'new_master_user=s' =>\$new_master_user,238);">'new_master_password=s' =>\$new_master_password,238);">'new_master_ssh_user=s' =>\$new_master_ssh_user,238);">'orig_master_ssh_port=i' =>\$orig_master_ssh_port,238);">'new_master_ssh_port=i' =>\$new_master_ssh_port,238);">sub current_time_us {
my ( $sec,$microsec ) = gettimeofday();
my $curdate = localtime($sec);
return $curdate . " " . sprintf( "%06d",$microsec);
sub sleep_until {
my $elapsed = tv_interval($_tstart);
if ( $_running_interval > $elapsed ) {
sleep( $_running_interval - $elapsed );
sub get_threads_util {
my $dbh = shift;
my $my_connection_id =shift;
my $running_time_threshold = shift;
my $type =shift;
$running_time_threshold = 0 unless ($running_time_threshold);
$type = 0 unless($type);
my @threads;
my $sth = $dbh->prepare("SHOW PROCESSLIST");
$sth->execute();
while ( my $ref = $sth->fetchrow_hashref() ) {
my $id = $ref->{Id};
my $user = $ref->{User};
my $host = $ref->{Host};
my $command =$ref->{Command};
my $state = $ref->{State};
my $query_time = $ref->{Time};
my $info = $ref->{Info};
$info =~ s/^\s*(.*?)\s*$/$1/ if defined($info);
next if ( $my_connection_id == $id );
next if ( defined($query_time) && $query_time < $running_time_threshold);
next if ( defined($command)&& $command eq "Binlog Dump" );
next if ( defined($user)&& $user eq "system user" );
next
if ( defined($command)
&& $command eq "Sleep"
&& defined($query_time)
&& $query_time >= 1);
if ( $type >= 1 ) {
next if ( defined($command) &&$command eq "Sleep" );
next if ( defined($command) &&$command eq "Connect" );
if ( $type >= 2 ) {
next if ( defined($info) && $info=~ m/^select/i );
next if ( defined($info) && $info=~ m/^show/i );
push @threads,$ref;
return @threads;
if ( $command eq "stop" ) {
## Gracefully killing connections on the current master
# 1. Set read_only= 1 on the new master
# 2. DROP USER so that no app user can establish new connections
# 3. Set read_only= 1 on the current master
# 4. Kill current queries
# * Any database access failure will result in script die.
my $exit_code = 1;
eval {
## Setting read_only=1 on the new master(to avoid accident)
my $new_master_handler = newMHA::DBHelper();
# args: hostname,port,user,password,raise_error(die_on_error)_or_not
$new_master_handler->connect($new_master_ip,238);"> $new_master_user,$new_master_password,1 );
print current_time_us() . " Setread_only on the new master.. ";
$new_master_handler->enable_read_only();
if ( $new_master_handler->is_read_only()) {
print "ok.\n";
else {
die "Failed!\n";
$new_master_handler->disconnect();
# Connecting to the orig master,die ifany database error happens
my $orig_master_handler = newMHA::DBHelper();
$orig_master_handler->connect($orig_master_ip,238);"> $orig_master_user,$orig_master_password,238);"> ## Drop application user so that nobodycan connect. Disabling per-session binlog beforehand
$orig_master_handler->disable_log_bin_local();
print current_time_us() . " Drppingapp user on the orig master..\n";
#FIXME_xxx_drop_app_user($orig_master_handler);
## Waiting for N * 100 milliseconds sothat current connections can exit
my $time_until_read_only = 15;
$_tstart = [gettimeofday];
my @threads = get_threads_util($orig_master_handler->{dbh},238);">$orig_master_handler->{connection_id} );
while ( $time_until_read_only > 0&& $#threads >= 0 ) {
if ( $time_until_read_only % 5 == 0 ) {
printf
"%s Waiting all running %dthreads are disconnected.. (max %d milliseconds)\n",238);"> current_time_us(),$#threads + 1,$time_until_read_only * 100;
if ( $#threads < 5 ) {
printData::Dumper->new( [$_] )->Indent(0)->Terse(1)->Dump ."\n"
foreach (@threads);
sleep_until();
$time_until_read_only--;
@threads = get_threads_util($orig_master_handler->{dbh},238);"> ## Setting read_only=1 on the currentmaster so that nobody(except SUPER) can write
print current_time_us() . " Setread_only=1 on the orig master.. ";
$orig_master_handler->enable_read_only();
if ($orig_master_handler->is_read_only() ) {
## Waiting for M * 100 milliseconds sothat current update queries can complete
my $time_until_kill_threads = 5;
while ( $time_until_kill_threads > 0&& $#threads >= 0 ) {
if ( $time_until_kill_threads % 5 == 0) {
"%s Waiting all running %dqueries are disconnected.. (max %d milliseconds)\n",$time_until_kill_threads * 100;
print Data::Dumper->new( [$_])->Indent(0)->Terse(1)->Dump . "\n"
$time_until_kill_threads--;
## Terminating all threads
print current_time_us() . " Killingall application threads..\n";
$orig_master_handler->kill_threads(@threads)if ( $#threads >= 0 );
print current_time_us() . "done.\n";
$orig_master_handler->enable_log_bin_local();
$orig_master_handler->disconnect();
## After finishing the script,MHAexecutes FLUSH TABLES WITH READ LOCK
`ssh -p$orig_master_ssh_port$orig_master_ssh_user\@$orig_master_host \" $ssh_stop_vip \"`;
if ($@) {
exit $exit_code;
## Activating master ip on the new master
# 1. Create app user with write privileges
# 2. Moving backup script if needed
# 3. Register new master's ip to the catalog database
# We don't return error eventhough activating updatable accounts/ip Failed so that we don't interruptslaves' recovery.
# If exit code is 0 or 10,MHAdoes not abort
my $exit_code = 10;
## Set read_only=0 on the new master
$new_master_handler->disable_log_bin_local();
print current_time_us() . " Setread_only=0 on the new master.\n";
$new_master_handler->disable_read_only();
## Creating an app user on the new master
print current_time_us() . " Creatingapp user on the new master..\n";
#FIXME_xxx_create_app_user($new_master_handler);
$new_master_handler->enable_log_bin_local();
## Update master ip on the catalogdatabase,etc
`ssh -p$new_master_ssh_port$new_master_ssh_user\@$new_master_host \" $ssh_start_vip \"`;
# do nothing
exit 0;
&usage();
exit 1;
"Usage:master_ip_online_change --command=start|stop|status --orig_master_host=host--orig_master_ip=ip --orig_master_port=port --new_master_host=host--new_master_ip=ip --new_master_port=port\n";
die;
在node94上检查MHA 的SSH是否配置无误：
masterha_check_ssh--conf=/etc/masterha/app1.cnf
如能出现"All SSH connection tests passedsuccessfully." 说明配置没问题
在node94上检查MHA 的主从复制是否配置无误：
masterha_check_repl--conf=/etc/masterha/app1.cnf
如提示"MysqL Replication Health is OK"说明配置没问题
在node94上前台启动MHA：
masterha_manager--conf=/etc/masterha/app1.cnf --ignore_last_failover 在前台启动监控
模拟node93master宕机、观察master的自动切换：
将node93的MysqL服务停掉，可以发现此时node94上开启的masterha_manager进程自动退出了，到其他节点去查看，可以发现主从切换了。
然后启动node93的MysqL，再次上线就不会自动变成master
【!!注意：直接将node93上线的话，集群中就出现了2个主节点，脑裂，masterha_manger也无法启动】，我们需要先手工将其改为从节点，操作如下：
在node93上，执行：
> change master to
master_host='10.1.20.94',
master_user='rpl',
master_password='rpl',
master_log_file='MysqL-bin.000003',
master_log_pos=1881; # 这里的位置，需要看下node94的/data/masterha/app1目录下的manager.log里面的内容找到具体的binlog位置。
> start slave;
> show slave status\G
将node93 重新加入集群后，我们在node94 manager节点上再次执行 masterha_manager--conf=/etc/masterha/app1.cnf --ignore_last_failover 发现启动不退出了。(验证过程中不要关闭这个窗口)
检查下当前主从的配置：
node94另外开一个xshell窗口，可以执行 masterha_check_repl--conf=/etc/masterha/app1.cnf
可以看出主节点、从节点的发生了变化:
查看是否masterha启动：
另外开一个xshell窗口，可以执行 masterha_check_status--conf=/etc/masterha/app1.cnf
如果需要停止masterha的话，不要用stop或者kill，要用下面的命令：
masterha_stop--conf=/etc/masterha/app1.cnf
手动切换主从的方法：
masterha_master_switch-h 查看帮助信息
masterha_master_switch--conf=/etc/masterha/app1.cnf --master_state=alive --new_master_host=10.1.20.93--new_master_port=3306 --orig_master_is_new_slave --running_updates_limit=10000
手工切换的时候需要注意2点:
1、执行手工切换的时候，需要先关掉老的master和即将提升为master的主机的 event scheduler，否则无法切换(set global event_scheduler = OFF;)
2、执行手工切换的时候，需要先关闭MHA的监控 masterha_stop--conf=/etc/masterha/app1.cnf )
3、执行手工切换脚本的时候，它会自动在原先的master上执行FLUSH TABLES WITH READ LOCK; 等切换完成后，再UNLOCK TABLES释放掉这个原有master的锁。
yuminstall perl-Mail-Sender
发送失败的话，可以查看/tmp/monitormail.log 找找失败的原因。
MHA异常的话：
可以查看日志路径：/data/masterha/app1/
masterha_manager 还有几个比较有用的启动参数：
--remove_dead_master_conf 该参数代表当发生主从切换后，老的主库的ip将会从配置文件中移除。
--manger_log 日志存放位置,想规范化管理日志可以加上
--ignore_last_failover 该参数代表忽略上次MHA触发切换产生的文件，默认情况下，MHA发生切换后会在日志目录，也就是上面我设置的/data产生app1.failover.complete文件，下次再次切换的时候如果发现该目录下存在该文件将不允许触发切换，除非在第一次切换后收到删除该文件，在缺省情况下，如果MHA检测到连续发生宕机，且两次宕机间隔不足8小时的话，则不会进行Failover，之所以这样限制是为了避免ping-pong效应。【如果我们需要强制切换的话，需要先移除这个文件app1.failover.complete】
原文链接：https://www.f2er.com/centos/377561.html