用户名: 密码: 忘记密码? 注册

heartbeat实现mysql ha

作者:  时间: 2010-09-10
环境说明:本环境由两台mysql 数据库和heartbeat 组成,一台的ip
为192.168.10.197,一台为192.168.10.198,对外提供服务的vip
为192.168.10.200
备注:heartbeat 本身是不能做到服务不可用自动切换的,所以用结合
额外的脚本才可以做到,本文中提到的moniter 脚本即为实现某个
mysql 服务不可用的时候自动切换的还可以自动报警
安装和配置过程分为如下几步:
第一部分:mysql 的安装配置
1 安装
1.1. 添加mysql 运行所需的用户和组
groupadd mysql
useradd -g mysql mysql
1.2. 解压安装
tar mysql-5.1.26-rc-linux-x86_64-glibc23.tar.gz
mv mysql-5.1.26-rc-linux-x86_64-glibc23 /usr/local/mysql/
chown -R mysql:mysql /usr/local/mysql
1.3. 复制主配置文件和启动脚本
cd /usr/local/mysql
cp support-files/my-medium.cnf /etc/my.cf
cp support-files/mysql.server /etc/init.d/mysqld
chmod +x /etc/init.d/mysqld
1.4. 初始化数据库
cd /usr/local/mysql
scripts/mysql_install_db --user=mysql
1.5. 更改数据目录的所有者和组
作者:曲宝全,转载标明出处!
2 / 14
chown -R mysql:mysql ./data
1.6. 注册mysql 为系统服务
chkconfig --add mysqld
chkconfig --levels 2345 mysqld on
1.7. 输出环境变量
#Vi /etc/profile(内容如下)
PATH=$PATH:/usr/local/mysql/bin
# source /etc/profile
1.8. 设置数据库密码
service mysqld start
mysqladmin password 123456(密码自己定这里只是个例子)
1.9. 开启root 远程访问权限
mysql –p123456
mysql> grant all on *.* to root@'%' identified by '123456';
Query OK, 0 rows affected (0.00 sec)
mysql> flush privileges;
mysql> quit
2 参数设置
2.1. 创建相关目录并设置权限
mkdir /usr/local/mysql/binlog/
chown –R mysql:mysql /usr/local/mysql/binlog/
touch /var/log/mysql.log
chown –R mysql:mysql /var/log/mysql.log
2.2. 添加同步复制的账号(主库和辅库的设置相同)
[root@master ~]# mysql -p
Enter password:(输入root 密码)
mysql> grant all on *.* to qiangao identified by '123456';(其中的账号和密码要
和配置文件中指定的相同)
mysql> flush privileges;
2.3. 修改主配置文件
作者:曲宝全,转载标明出处!
3 / 14
注意:其中ip 要互相指向对方的ip 按实际情况来设定,用户名和密码要和上文的用户名密
码保持一致
主库设置
vi /etc/my.cnf(用如下内容覆盖原有内容)
################################################
##########
[client]
port = 3306
socket = /tmp/mysql.sock
[mysqld]
################### auto_increment
###########################
auto_increment_offset = 1
auto_increment_increment = 2
############## other options ##############
default-character-set = utf8
default-storage-engine = InnoDB
default-table-type = INNODB
max_connections = 800
port = 3306
socket = /tmp/mysql.sock
skip-locking
########## MyISAM options #################
myisam_max_sort_file_size = 10G
myisam_max_extra_sort_file_size = 10G
myisam_sort_buffer_size = 10M
myisam_repair_threads = 1
################ select cache options ##################
read_buffer_size = 2M
read_rnd_buffer_size = 16M
bulk_insert_buffer_size = 5M
作者:曲宝全,转载标明出处!
4 / 14
max_allowed_packet = 1M
table_cache = 2048
query_cache_size = 32M
query_cache_limit = 2M
sort_buffer_size = 8M
join_buffer_size = 8M
thread_concurrency = 8
################ index cache options ##################
key_buffer_size = 32M
key_buffer_size = 32M
################# master #########################
server-id = 1
log-bin=/usr/local/mysql/binlog/master-bin
binlog_format=mixed
relay-log=/usr/local/mysql/binlog/mysqld-relay-bin
##################### slave
##########################
relay-log=/usr/local/mysql/binlog/slave-relay-bin
master-host=192.168.10.198
master-user=qiangao
master-password=123456
master-connect-retry=10
############## log ##################
log-error=/var/log/mysql.log
######### INNODB #########
innodb_file_per_table
#+++++++ log ++++++++#
innodb_log_buffer_size = 10M
innodb_mirrored_log_groups = 1
innodb_log_files_in_group = 3
作者:曲宝全,转载标明出处!
5 / 14
innodb_log_file_size = 50M
innodb_flush_log_at_trx_commit = 0
#innodb_log_archive = 0
#+++++++ System buffer +++++++#
innodb_buffer_pool_size = 1024M
innodb_additional_mem_pool_size = 40M
#+++++++ Other +++++++#
innodb_file_io_threads = 4
innodb_lock_wait_timeout = 5
innodb_force_recovery = 0
innodb_fast_shutdown = 1
innodb_thread_concurrency = 8
innodb_lock_wait_timeout = 50
transaction-isolation = READ-COMMITTED
[mysqldump]
quick
max_allowed_packet = 16M
[mysql]
no-auto-rehash
# Remove the next comment character if you are not familiar with SQL
#safe-updates
[isamchk]
key_buffer = 128M
sort_buffer_size = 128M
read_buffer = 2M
write_buffer = 2M
[myisamchk]
key_buffer = 128M
sort_buffer_size = 128M
read_buffer = 2M
作者:曲宝全,转载标明出处!
6 / 14
write_buffer = 2M
[mysqlhotcopy]
interactive-timeout
辅库设置
vi /etc/my.cnf(用如下内容覆盖原有内容,)
################################################
##########
[client]
port = 3306
socket = /tmp/mysql.sock
[mysqld]
################### auto_increment
###########################
auto_increment_offset = 1
auto_increment_increment = 2
############## other options ##############
default-character-set = utf8
default-storage-engine = InnoDB
default-table-type = INNODB
max_connections = 800
port = 3306
socket = /tmp/mysql.sock
skip-locking
########## MyISAM options #################
myisam_max_sort_file_size = 10G
myisam_max_extra_sort_file_size = 10G
myisam_sort_buffer_size = 10M
myisam_repair_threads = 1
################ select cache options ##################
read_buffer_size = 2M
read_rnd_buffer_size = 16M
作者:曲宝全,转载标明出处!
7 / 14
bulk_insert_buffer_size = 5M
max_allowed_packet = 1M
table_cache = 2048
query_cache_size = 32M
query_cache_limit = 2M
sort_buffer_size = 8M
join_buffer_size = 8M
thread_concurrency = 8
################ index cache options ##################
key_buffer_size = 32M
key_buffer_size = 32M
################# master #########################
server-id = 2
log-bin=/usr/local/mysql/binlog/master-bin
binlog_format=mixed
relay-log=/usr/local/mysql/binlog/mysqld-relay-bin
##################### slave
##########################
relay-log=/usr/local/mysql/binlog/slave-relay-bin
master-host=192.168.10.197
master-user=qiangao
master-password=123456
master-connect-retry=10
############## log ##################
log-error=/var/log/mysql/error.log
######### INNODB #########
innodb_file_per_table
#+++++++ log ++++++++#
innodb_log_buffer_size = 10M
innodb_mirrored_log_groups = 1
作者:曲宝全,转载标明出处!
8 / 14
innodb_log_files_in_group = 3
innodb_log_file_size = 50M
innodb_flush_log_at_trx_commit = 0
#innodb_log_archive = 0
#+++++++ System buffer +++++++#
innodb_buffer_pool_size = 1024M
innodb_additional_mem_pool_size = 40M
#+++++++ Other +++++++#
innodb_file_io_threads = 4
innodb_lock_wait_timeout = 5
innodb_force_recovery = 0
innodb_fast_shutdown = 1
innodb_thread_concurrency = 8
innodb_lock_wait_timeout = 50
transaction-isolation = READ-COMMITTED
[mysqldump]
quick
max_allowed_packet = 16M
[mysql]
no-auto-rehash
# Remove the next comment character if you are not familiar with SQL
#safe-updates
[isamchk]
key_buffer = 128M
sort_buffer_size = 128M
read_buffer = 2M
write_buffer = 2M
[myisamchk]
key_buffer = 128M
sort_buffer_size = 128M
作者:曲宝全,转载标明出处!
9 / 14
read_buffer = 2M
write_buffer = 2M
[mysqlhotcopy]
interactive-timeout
2.4. 重新启动服务
service mysqld restart
3 防火墙是设置
注意主库和辅库的防火墙设置是一样的
iptables -I INPUT -p tcp --dport 3306 -j ACCEPT
service iptables save
第二部分:heartbeat 的安装和配置
两台主机硬件环境(不必完全一致):
eth0: 对外IP
eth1: 对内IP(HA 专用)
两台主机的eht1 使用双机对联线直接连接。
【二】安装前网络环境设定:
================================================
========
node1: 主机名:master.qiangao.com
eth0: 192.168.10.197 //对外IP 地址
eth1: 172.16.1.3 //HA 心跳使用地址
---------------------------
node2: 主机名:slave.qiangao.com
eth0: 192.168.10.198 //对外IP 地址
eth1: 172.16.1.4 //HA 心跳使用地址
特别注意要检查以下几个文件:
/etc/hosts
/etc/host.conf
/etc/resolv.conf
/etc/sysconfig/network
/etc/sysconfig/network-scripts/ifcfg-eth0
/etc/sysconfig/network-scripts/ifcfg-eth1
/etc/nsswitch.conf
#vi /etc/hosts
node1 的hosts 内容如下:
127.0.0.1 master.qiangao.com master localhost.localdomain localhost
192.168.10.197 master.qiangao.com
作者:曲宝全,转载标明出处!
10 / 14
192.168.10.198 slave.qiangao.com
::1 localhost6.localdomain6 localhost6
---------------------------
node2 的hosts 内容如下:
127.0.0.1 slave.qiangao.com slave localhost.localdomain localhost
192.168.10.197 master.qiangao.com
192.168.10.198 slave.qiangao.com
::1 localhost6.localdomain6 localhost6
#cat /etc/host.conf
order hosts,bind
#cat /etc/resolv.conf
nameserver 202.96.134.133 //DNS 地址
#cat /etc/sysconfig/network
NETWORKING=yes
HOSTNAME= master.qiangao.com //主机名
GATEWAY="192.168.10.1" //网关
GATEWAY="eth0" //网关使用网卡
ONBOOT=YES //启动时加载
FORWARD_IPV4="yes" //只允许IPV4
---------------------------
#cat /etc/sysconfig/network-scripts/ifcfg-eth0
DEVICE=eth0
ONBOOT=yes
BOOTPROTO=static
IPADDR=192.168.10.197
NETMASK=255.255.255.0
GATEWAY=192.168.10.1
TYPE=Ethernet
IPV6INIT=no
---------------------------
#cat /etc/sysconfig/network-scripts/ifcfg-eth1
DEVICE=eth1
ONBOOT=yes
BOOTPROTO=none
IPADDR=172.16.1.3
NETMASK=255.255.0.0
TYPE=Ethernet
[node1] 与 [node2] 在上面的配置中,除了
/etc/hosts
/etc/sysconfig/network
/etc/sysconfig/network-scripts/ifcfg-eth0
/etc/sysconfig/network-scripts/ifcfg-eth1
要各自修改外,其他一致。
配置完成后,试试在各自主机上ping 对方的主机名,应该可以ping 通:
作者:曲宝全,转载标明出处!
11 / 14
[root@master ~]# ping slave.qiangao.com
PING slave.qiangao.com (192.168.10.198) 56(84) bytes of data.
64 bytes from slave.qiangao.com (192.168.10.198): icmp_seq=1 ttl=64
time=0.136 ms
【三】安装HA 与HA 依赖包
================================================
===========
yum install heartbeat
yum install ipvsadm
yum install libnet
【四】 配置 HA 的各配置文件
首先复制配置文件到/etc 目录
cd /usr/share/doc/heartbeat-2.1.3
cp ha.cf haresources /etc/ha.d/
cp authkeys /etc/ha.d/
================================================
===========
配置心跳的加密方式:authkeys
如果使用双机对联线(双绞线),可以配置如下:
#vi /etc/hc.d/authkeys
auth 1
1 crc
存盘退出,然后
#chmod 600 authkeys
================================================
===========
配置心跳的监控:haresources
创建检测脚本(本处如果没有个监测脚本hearbeat 无法启动)
vi /etc/init.d/test(内容如下)
#!/bin/bash
echo "" $>/dev/null
chmod 777 /etc/init.d/test
================================================
===========
#vi /etc/ha.d/haresources(配置资源文件)
各主机这部分应完全相同。
master.qiangao.com 192.168.10.200 test
指定 master.qiangao.com 调用预先写好的一个测试脚本,系统附加一个虚拟IP
192.168.10.200 给eth0:0
这里如果master.qiangao.com 宕机后slave.qiangao.com 可以新分配IP
192.168.10.200
================================================
===========
配置心跳的配置文件:ha.cf
作者:曲宝全,转载标明出处!
12 / 14
#vi /etc/ha.d/ha.cf
logfile /var/log/ha_log/ha-log.log ## ha 的日志文件记录位置。如没有
该目录,则需要手动添加
bcast eth1 ##使用eht1 做心跳监测
keepalive 2 ##设定心跳(监测)时间时间为2 秒
warntime 4####警告时间
deadtime 6#########确定服务以死的时间
initdead 30
hopfudge 1
udpport 694 ##使用udp 端口694 进行心跳监测
auto_failback on
node master.qiangao.com ##节点1,必须要与 uname -n 指令得到的结果一致。
node slave.qiangao.com ##节点2
ping 172.16.1.04 ##通过ping 对方来监测心跳是否正常。
respawn hacluster /usr/lib64/heartbeat/ipfail(因操作系统而异)
apiauth ipfail gid=haclient uid=hacluster
debugfile /var/logs/ha-debug.log
---------------------------
【五】 HA 服务的启动、关闭
[root@master ha.d]# chkconfig --add heartbeat
[root@master ha.d]# chkconfig --levels 2345 heartbeat on
启动HA: service heartbeat start
关闭HA; service heartbeat stop
【六】 防火墙设置
================================================
====
heartbeat 默认使用udp 694 端口进行心跳监测。 如果系统有使用iptables 做
防火墙,应记住把这个端口打开。
#vi /etc/sysconfig/iptables
加入以下内容(互相指向对方ip)
-A RH-Firewall-1-INPUT -p udp -m udp --dport 694 -d 172.16.1.4 -j ACCEPT
意思是udp 694 端口对 对方的心跳网卡地址 172.16.1.4 开放。
#service iptables restart
重新加载iptables。
第三部分:监控脚本
cat /usr/local/mysql/bin/moniter.sh
#!/bin/bash
mysql_path=/usr/local/mysql/bin/
user="root"
password="123456"
email="qubq@qian-gao.com"
logfile=/var/log/moniter.log
作者:曲宝全,转载标明出处!
13 / 14
date=`(date +%y-%m-%d--%H:%M:%S)`
sleeptime=30
ip=$(/sbin/ifconfig | grep "inet addr" | grep -v "127.0.0.1" | awk '{print $2;}' |
awk -F':' '{print $2;}' | head -1)
Slave_IO_Running=$(mysql -u$user -p$password -e 'show slave status\G' |
grep "Slave_IO_Running" | awk '{print $2}')
Slave_SQL_Running=$(mysql -u$user -p$password -e 'show slave status\G' |
grep "Slave_SQL_Running" | awk '{print $2}')
echo "plese fix the server of $ip error now!" >$mysql_path/letter
letter=$mysql_path/letter
mysql -p$password -e "use test;"
if [[ $? != 0 ]]
then
mail -s "{$ip}_database connect lost the srcprits fix it now "
$email<$letter
killall -9 heartbeat
killall -9 mysqld
/etc/init.d/mysqld start
sleep $sleeptime
mysql -p$password -e "use test;"
if [ $? == 0 ]
then
echo
"==============>$date<=====================">>$logfile
mail -s "{$ip}_database up now " $email
sleep $sleeptime
service heartbeat start
sleep $sleeptime
netstat -an |grep udp |grep 694
if [ $? == 0 ]
then
echo "complete!">>$logfile
else
mail -s "{$ip} heartbeat can't to up please fix it !"
$email<$letter
fi
else
mail -s "{$ip}_database cant't to up plese fix it" $email
fi
else
if [ "$Slave_IO_Running" = "Yes" -a "$Slave_SQL_Running" =
"Yes" ]
then
echo "Slave is running!" >/dev/null
作者:曲宝全,转载标明出处!
14 / 14
else
echo
"==============>$date<=====================">>$logfile
echo "Slave is not running!" >> $logfile
/bin/mail -s "{$ip}_replicate error please fix it "
$email<$letter
fi
fi
然后设置自动化任务每2分钟检测一次就可以了,现在就可以实现双机热备了!