CoroSync+Drbd+MySQL如何實(shí)現(xiàn)MySQL的高可用集群

這篇文章主要介紹CoroSync+Drbd+MySQL如何實(shí)現(xiàn)MySQL的高可用集群，文中介紹的非常詳細(xì)，具有一定的參考價(jià)值，感興趣的小伙伴們一定要看完！

節(jié)點(diǎn)規(guī)劃：

node1.huhu.com172.16.100.103

node2.huhu.com172.16.100.104

資源名稱規(guī)劃：

資源名稱：可以是除了空白字符外的任意ACSII碼字符

DRBD設(shè)備：在雙節(jié)點(diǎn)上，此DRBD設(shè)備文件，一般為/dev/drbdN，主設(shè)備號147

磁盤：在雙方節(jié)點(diǎn)上，各自提供存儲設(shè)備

網(wǎng)絡(luò)配置：雙方數(shù)據(jù)同步所使用的網(wǎng)絡(luò)屬性

DRBD從Linux內(nèi)核2.6.33起已經(jīng)整合進(jìn)內(nèi)核

1.配置雙擊互信（基于秘鑰認(rèn)證），HOSTS文件，時(shí)間同步

1）所有節(jié)點(diǎn)的主機(jī)名稱和對應(yīng)的IP地址解析服務(wù)可以正常工作，且每個(gè)節(jié)點(diǎn)的主機(jī)名稱需要跟"uname -n“命令的結(jié)果保持一致；因此，需要保證兩個(gè)節(jié)點(diǎn)上的/etc/hosts文件均為下面的內(nèi)容：

172.16.100.103node1.huhu.com node1   172.16.100.104node2.huhu.com node2

Node1:

#sed -i 's@\(HOSTNAME=\).*@\1node1.huhu.com@g' /etc/sysconfig/network   #hostname node1.huhu.com

Node2：

#sed -i 's@\(HOSTNAME=\).*@\1node2.huhu.com@g' /etc/sysconfig/network   #hostname node2.huhu.com

2）設(shè)定兩個(gè)節(jié)點(diǎn)可以基于密鑰進(jìn)行ssh通信，這可以通過類似如下的命令實(shí)現(xiàn)：

#yum install openssh-clients

Node1:

#ssh-keygen -t rsa   #ssh-copy-id -i ~/.ssh/id_rsa.pub root@node2

Node2:

#ssh-keygen -t rsa   #ssh-copy-id -i ~/.ssh/id_rsa.pub root@node1

配置時(shí)間同步：

*/5* * * * root /usr/sbin/ntpdate ntp.api.bz & > /dev/null

2.創(chuàng)建和配置DRBD

Node1上執(zhí)行：

#rpm -Uvh http://www.elrepo.org/elrepo-release-6-6.el6.elrepo.noarch.rpm   #ssh node2 'rpm -Uvh http://www.elrepo.org/elrepo-release-6-6.el6.elrepo.noarch.rpm'  #yum update -y   #ssh node2 'yum update -y'   #yum install drbd84-utils kmod-drbd84 -y   #ssh node2 'yum install drbd84-utils kmod-drbd84 -y'

加載模塊到內(nèi)核：

#/sbin/modprobe drbd   #ssh node2 '/sbin/modprobe drbd'

DRBD的配置文件：

/etc/drbd.conf  /etc/drbd.d/global_common.conf  /etc/drbd.d/resource.d/  #yum -y install parted   #ssh node2 'yum -y install parted'   #fdisk /dev/sdb

n新建分區(qū)

p主分區(qū)

1分區(qū)號，兩車回車按照默認(rèn)大小選擇

wq保存退出

#partprobe /dev/sdb1

資源規(guī)劃：

資源名稱：mydrbd

DRBD設(shè)備：/dev/drbd0

磁盤：/dev/sdb1

網(wǎng)絡(luò)配置：100M

#cat /etc/drbd.d/global_common.conf | grep -v "#"

#cat /etc/drbd.d/global_common.conf | grep -v "#"   global{   usage-countyes;   }  common{   handlers{   pri-on-incon-degr"/usr/lib/drbd/notify-pri-on-incon-degr.sh;/usr/lib/drbd/notify-emergency-reboot.sh; echo b > /proc/sysrq-trigger ;reboot -f";   pri-lost-after-sb"/usr/lib/drbd/notify-pri-lost-after-sb.sh;/usr/lib/drbd/notify-emergency-reboot.sh; echo b > /proc/sysrq-trigger ;reboot -f";   local-io-error"/usr/lib/drbd/notify-io-error.sh; /usr/lib/drbd/notify-emergency-shutdown.sh;echo o > /proc/sysrq-trigger ; halt -f";   }  startup{   }  options{   }  disk{   on-io-errordetach;   }  net{   cram-hmac-alg "sha1";   shared-secret "1q2w3e4r5t6y";   }  syncer{   rate 200M;   }  }  #cat mydrbd.res   resourcemydrbd {   device/dev/drbd0;   disk/dev/sdb1;   meta-diskinternal;   onnode1.huhu.com {   address172.16.100.103:7789;   }  onnode1.huhu.com {   address172.16.100.104:7789;   }  }

復(fù)制配置文件到node2節(jié)點(diǎn)

scp-r /etc/drbd.* node2:/etc/

在兩個(gè)節(jié)點(diǎn)上，初始化已定義的資源，并啟動服務(wù)

#drbdadm create-md mydrbd

#ssh node2 'drbdadm create-md mydrbd'   #/etc/init.d/drbd start   #ssh node2 '/etc/init.d/drbd start'

查看DRBD設(shè)備的狀態(tài)：

#cat /proc/drbd   version:8.4.4 (api:1/proto:86-101)   GIT-hash:599f286440bd633d15d5ff985204aff4bccffadd build by phil@Build64R6, 2013-10-1415:33:06   0:cs:Connected ro:Secondary/Secondaryds:Inconsistent/Inconsistent C r-----   ns:0nr:0 dw:0 dr:0 al:0 bm:0 lo:0 pe:0 ua:0 ap:0 ep:1 wo:f oos:2096348

目前兩個(gè)節(jié)點(diǎn)都處于secondary狀態(tài)，手動讓其node1成為主節(jié)點(diǎn)：

#drbdadm -- --overwrite-data-of-peer primary mydrbd   #cat /proc/drbd   version:8.4.4 (api:1/proto:86-101)   GIT-hash:599f286440bd633d15d5ff985204aff4bccffadd build by phil@Build64R6, 2013-10-1415:33:06   0:cs:Connected ro:Primary/Secondary ds:UpToDate/UpToDateC r-----   ns:2096348nr:0 dw:0 dr:2097012 al:0 bm:128 lo:0 pe:0 ua:0 ap:0 ep:1 wo:f oos:0

在primary節(jié)點(diǎn)上進(jìn)行格式化drbd分區(qū)然后掛載

#mke2fs -j /dev/drbd0

#mkdir /mydata

#mount /dev/drbd0 /mydata/

#cp /etc/inittab /mydata/

#ls -lh /mydata/

total20K

-rw-r--r--.1 root root 884 Jul 8 17:24 inittab

drwx------.2 root root 16K Jul 8 17:23 lost+found

到此drbd分區(qū)已經(jīng)可以正常使用了

DRBD分區(qū)的主備切換

primary節(jié)點(diǎn)上執(zhí)行：

#umount /mydata/

# drbdadm secondary mydrbd

#drbd-overview

0:mydrbd/0Connected Secondary/Secondary UpToDate/UpToDateC r-----

secondary上執(zhí)行

#drbd-overview 確保兩者的狀態(tài)都有secondary狀態(tài)

#drbdadm primary mydrbd

#mkdir -p /mydata

# mount /dev/drbd0 /mydata/

#ls -lh /mydata/

total20K

-rw-r--r--.1 root root 884 Jul 8 17:24 inittab

drwx------.2 root root 16K Jul 8 17:23 lost+found

#drbd-overview

0:mydrbd/0Connected Primary/Secondary UpToDate/UpToDate Cr----- /mydata ext3 2.0G 36M 1.9G 2%

狀態(tài)已經(jīng)改變，primary /secondary

3.配置coresync服務(wù)

各個(gè)節(jié)點(diǎn)上停止掉drbd服務(wù)，并且關(guān)閉開機(jī)啟動

#/etc/init.d/drbd stop

#ssh node2 '/etc/init.d/drbd stop'

#chkconfig drbd off

#ssh node2 'chkconfig drbd off'

#chkconfig --list | grep drbd

#ssh node2 'chkconfig --list | grep drbd'

drbd0:off 1:off 2:off 3:off 4:off 5:off 6:off

安裝corosync

#yum install libibverbs librdmacm lm_sensors libtool-ltdl openhpi-libs openhpiperl-TimeDate

#yum install corosync pacemaker

#ssh node2 '# yum install libibverbs librdmacm lm_sensors libtool-ltdlopenhpi-libs openhpi perl-TimeDate'

wgethttp://ftp5.gwdg.de/pub/opensuse/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-6/x86_64/crmsh-2.1-1.1.x86_64.rpm&& wgethttp://ftp5.gwdg.de/pub/opensuse/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-6/x86_64/pssh-2.3.1-4.1.x86_64.rpm

#ssh node2 'wget http://ftp5.gwdg.de/pub/opensuse/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-6/x86_64/crmsh-2.1-1.1.x86_64.rpm&& wgethttp://ftp5.gwdg.de/pub/opensuse/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-6/x86_64/pssh-2.3.1-4.1.x86_64.rpm'

yum--nogpgcheck localinstall crmsh-2.1-1.1.x86_64.rpm pssh-2.3.1-4.1.x86_64.rpm

如果安裝失敗，請?zhí)砑右韵略?/p>

#vim /etc/yum.repos.d/ha-clustering.repo

[haclustering]

name=HAClustering

baseurl=http://download.opensuse.org/repositories/network:/ha-clustering:/Stable/CentOS_CentOS-6/

enabled=1

gpgcheck=0

#yum --nogpgcheck localinstall crmsh-2.1-1.1.x86_64.rpmpssh-2.3.1-4.1.x86_64.rpm

node2上同樣執(zhí)行以上

配置corosync

#cd /etc/corosync/

#cp corosync.conf.example corosync.conf

#cat corosync.conf | grep -v "^#" | sed -e'/^$/d'

compatibility: whitetank

totem {

version: 2

secauth: on

threads: 2

interface {

ringnumber: 0

bindnetaddr: 172.16.100.0

mcastaddr: 226.94.8.9

mcastport: 5405

ttl: 1

}

}

logging {

fileline: off

to_stderr: no

to_logfile: yes

to_syslog: no

logfile: /var/log/cluster/corosync.log

debug: off

timestamp: on

logger_subsys {

subsys: AMF

debug: off

}

}

service {

ver: 0

name: pacemaker

# use_mgmtd: yes

}

aisexec {

user: root

group: root

}

amf {

mode: disabled

}

生成秘鑰

#corosync-keygen

#scp -p authkey corosync.conf node2:/etc/corosync/

創(chuàng)建日志文件目錄

#mkdir -p /var/log/cluster/ -pv

# ssh node2 'mkdir -p /var/log/cluster/ -pv'

啟動corosync服務(wù)

#service corosync start

#ssh node2 'service corosync start'

檢查corosync引擎是否已經(jīng)啟動

#grep -e "Corosync Cluster Engine" -e "configuration file"/var/log/cluster/corosync.log

Jul09 10:28:14 corosync [MAIN ] Corosync Cluster Engine ('1.4.1'): started andready to provide service.

Jul09 10:28:14 corosync [MAIN ] Successfully readmain configuration file '/etc/corosync/corosync.conf'.

查看節(jié)點(diǎn)成員之間通信是否正常

#grep TOTEM /var/log/cluster/corosync.log

Jul09 10:28:14 corosync [TOTEM ] Initializing transport (UDP/IP Multicast).

Jul09 10:28:14 corosync [TOTEM ] Initializing transmit/receive security:libtomcrypt SOBER128/SHA1HMAC (mode 0).

Jul09 10:28:14 corosync [TOTEM ] The network interface[172.16.100.103] is now up.

Jul09 10:28:14 corosync [TOTEM ] A processor joined or left the membership and anew membership was formed.

Jul09 10:28:29 corosync [TOTEM ] A processor joined or left the membership and anew membership was formed.

檢查pacemaker啟動是否正常

#grep pcmk_startup /var/log/cluster/corosync.log

Jul09 10:28:14 corosync [pcmk ] info: pcmk_startup: CRM: Initialized

Jul09 10:28:14 corosync [pcmk ] Logging: Initialized pcmk_startup

Jul09 10:28:14 corosync [pcmk ] info: pcmk_startup:Maximum core file size is: 18446744073709551615

Jul09 10:28:14 corosync [pcmk ] info: pcmk_startup: Service: 9

Jul09 10:28:14 corosync [pcmk ] info: pcmk_startup: Local hostname: node1.huhu.com

查看錯(cuò)誤信息

#grep ERROR /var/log/cluster/corosync.log | grep -v unpack_resources

Jul09 10:28:14 corosync [pcmk ] ERROR: process_ais_conf: You have configured acluster using the Pacemaker plugin for Corosync. The plugin is not supported inthis environment and will be removed very soon.

Jul09 10:28:14 corosync [pcmk ] ERROR: process_ais_conf: Please see Chapter 8 of'Clusters from Scratch' (http://www.clusterlabs.org/doc) for details on usingPacemaker with CMAN

Jul09 10:28:35 [1373] node1.huhu.com pengine: notice: process_pe_message:Configuration ERRORs found during PE processing. Please run "crm_verify-L" to identify issues.

注意：這里因?yàn)闆]有使用stonith設(shè)備，因此錯(cuò)誤可以忽略

#crm status

Lastupdated: Wed Jul 9 10:49:53 2014

Lastchange: Wed Jul 9 10:19:07 2014 via crmd on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node1.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

0Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

以上說明corosync配置啟動正常。

關(guān)閉stonith設(shè)備，并驗(yàn)證提交

crm(live)#configure

crm(live)configure#property stonith-enabled=false

crm(live)configure#verify

crm(live)configure#commit

關(guān)閉不具備法定票數(shù)的時(shí)候，不能關(guān)閉集群服務(wù)

crm(live)configure# property no-quorum-policy=ignore

crm(live)configure#verify

crm(live)configure#commit

配置資源粘性，更傾向于當(dāng)前節(jié)點(diǎn)

crm(live)configure#rsc_defaults resource-stickiness=100

crm(live)configure#verify

crm(live)configure#commit

查看當(dāng)前的配置

crm(live)configure#show

nodenode1.huhu.com

nodenode2.huhu.com

propertycib-bootstrap-options: \

dc-version=1.1.10-14.el6_5.3-368c726\

cluster-infrastructure="classicopenais (with plugin)" \

expected-quorum-votes=2\

stonith-enabled=false\

no-quorum-policy=ignore

rsc_defaultsrsc-options: \

resource-stickiness=100

crm(live)configure#

查看drbd的資源代理

crm(live)configure#cd ..

crm(live)#ra

crm(live)ra#providers drbd

linbit

注意：這里只有l(wèi)inbit沒有beartbeat，corosync1.4以前的版本有heartbeat。

查看原數(shù)據(jù)

crm(live)ra#meta ocf:linbit:drbd

定義資源：

crm(live)configure#primitive mysql_drbd ocf:linbit:drbd paramsdrbd_resource=mydrbd op start timeout=240 op stop timeout=100 op monitorrole=Master interval=50s timeout=30s op monitor role=Slave interval=60s timeout=30s

定義群集資源：

crm(live)configure#master MS_mysql_drbd mysql_drbd metamaster-max="1" master-node-max="1" clone-max="2"clone-node-max="1" notify="true"

crm(live)configure#show mysql_drbd

primitivemysql_drbd ocf:linbit:drbd \

paramsdrbd_resource=mydrbd \

opstart timeout=240 interval=0 \

opstop timeout=100 interval=0 \

opmonitor role=Master interval=50s timeout=30s \

opmonitor role=Slave interval=60s timeout=30s

crm(live)configure#show MS_mysql_drbd

msMS_mysql_drbd mysql_drbd \

metamaster-max=1 master-node-max=1 clone-max=2 clone-node-max=1 notify=true

crm(live)configure#verify

crm(live)configure#commit

crm(live)configure#cd

crm(live)#status

Lastupdated: Wed Jul 9 11:54:30 2014

Lastchange: Wed Jul 9 11:54:17 2014 via cibadmin on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

2Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

Master/Slave Set: MS_mysql_drbd [mysql_drbd]

Masters: [ node1.huhu.com ]

Slaves: [ node2.huhu.com ]

crm(live)#

主從資源已經(jīng)定義完成

[root@node1corosync]# drbd-overview

0:mydrbd/0Connected Primary/Secondary UpToDate/UpToDate C r-----

[root@node1corosync]#

此時(shí)當(dāng)前節(jié)點(diǎn)已經(jīng)成為主資源了

手動做一次主從切換：

#crm node standby

#crm status

Lastupdated: Wed Jul 9 12:01:44 2014

Lastchange: Wed Jul 9 12:01:29 2014 via crm_attribute on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

2Resources configured

Nodenode1.huhu.com: standby

Online:[ node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters:[ node2.huhu.com ]

Stopped:[ node1.huhu.com ]

#crm node online

#crm status

Lastupdated: Wed Jul 9 12:02:46 2014

Lastchange: Wed Jul 9 12:02:43 2014 via crm_attribute on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

2Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters:[ node2.huhu.com ]

Slaves:[ node1.huhu.com ]

#drbd-overview

0:mydrbd/0Connected Secondary/Primary UpToDate/UpToDate C r-----

[root@node1corosync]#

當(dāng)前節(jié)點(diǎn)就切換為從節(jié)點(diǎn)了

此時(shí)保證了資源可以主從切換，但是文件系統(tǒng)是沒有掛載

因此必須定義文件系統(tǒng)

crm(live)configure#primitive mystore ocf:heartbeat:Filesystem paramsdevice=/dev/drbd0 directory=/mydata fstype=ext3 op start timeout=60 op stoptimeout=60

crm(live)configure#verify

注意：這里千萬不要提交，因?yàn)楸仨毐ＷC文件系統(tǒng)跟主節(jié)點(diǎn)在一起，定義排列約束

crm(live)configure#colocation mystore_with_MS_mysql_drbd inf: mystoreMS_mysql_drbd:Master

定義存儲資源必須和資源的主節(jié)點(diǎn)在一起

crm(live)configure#order mystore_after_MS_mysql_drbd mandatory:MS_mysql_drbd:promote mystore:start

定義存儲資源必須在主節(jié)點(diǎn)啟動后進(jìn)行掛載

crm(live)configure#verify

crm(live)configure#commit

crm(live)configure#cd ..

crm(live)#status

Lastupdated: Wed Jul 9 12:25:25 2014

Lastchange: Wed Jul 9 12:22:30 2014 via cibadmin on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

3Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters:[ node1.huhu.com ]

Slaves:[ node2.huhu.com ]

mystore(ocf::heartbeat:Filesystem): Started node1.huhu.com

crm(live)#

可以看到Master在node1上，mystore就啟動在node1上

[root@node1~]# ls -lh /mydata/

total20K

-rw-r--r--.1 root root 884 Jul 8 17:24 inittab

drwx------2 root root 16K Jul 8 17:23 lost+found

[root@node1~]#

手動模擬一次切換

[root@node1corosync]# crm node standby

[root@node1corosync]# crm status

Lastupdated: Wed Jul 9 12:28:55 2014

Lastchange: Wed Jul 9 12:28:49 2014 via crm_attribute on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

3Resources configured

Nodenode1.huhu.com: standby

Online:[ node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters:[ node2.huhu.com ]

Stopped:[ node1.huhu.com ]

mystore(ocf::heartbeat:Filesystem): Started node2.huhu.com

[root@node1corosync]#

[root@node2~]# ls -lh /mydata/

total20K

-rw-r--r--.1 root root 884 Jul 8 17:24 inittab

drwx------2 root root 16K Jul 8 17:23 lost+found

Youhave new mail in /var/spool/mail/root

[root@node2~]#

這樣就切換到node2節(jié)點(diǎn)上。

4.配置MySQL結(jié)合DRBD和corosync

分別在node1節(jié)點(diǎn)上創(chuàng)建MySQL用戶和組

#groupadd -g 3306 mysql

#useradd -u 3306 -g mysql -s /sbin/nologin -M mysql

#id mysql

uid=3306(mysql)gid=3306(mysql) groups=3306(mysql)

#ssh node2 'groupadd -g 3306 mysql'

#ssh node2 'useradd -u 3306 -g mysql -s /sbin/nologin -M mysql'

#wgethttp://cdn.mysql.com/Downloads/MySQL-5.5/mysql-5.5.38-linux2.6-x86_64.tar.gz

#tar zxvf mysql-5.5.38-linux2.6-x86_64.tar.gz -C /usr/local/

#cd /usr/local/

#ln -s mysql-5.5.38-linux2.6-x86_64/ mysql

#cd mysql

#chown root:mysql -R .

#cp support-files/my-huge.cnf /etc/my.cnf

#cp support-files/mysql.server /etc/init.d/mysqld

#[ -x /etc/init.d/mysqld ] && echo "ok" || echo "NO"

確保當(dāng)前在主節(jié)點(diǎn)上操作

#drbd-overview

0:mydrbd/0Connected Primary/Secondary UpToDate/UpToDate Cr----- /mydata ext3 2.0G 36M 1.9G 2%

#mkdir -p /mydata/data

#chown -R mysql:mysql /mydata/data/

#scripts/mysql_install_db --user=mysql--datadir=/mydata/data

vim/etc/my.cnf

datadir=/mydata/data

#chkconfig --add mysqld

#chkconfig mysqld off

#service mysqld start

確保啟動OK

#/usr/local/mysql/bin/mysql -uroot -e "CREATE DATABASE mydb"

[root@node1mysql]# /usr/local/mysql/bin/mysql -uroot -e "SHOW DATABASES"

+--------------------+

|Database |

+--------------------+

|information_schema |

|mydb |

|mysql |

|performance_schema |

|test |

+--------------------

#service mysqld stop

#chkconfig --list | grep 3:off | grep mysql

mysqld0:off 1:off 2:off 3:off 4:off 5:off 6:off

[root@node1mysql]#

將存儲資源切換到node2上，在node2MySQL配置好

#crm node standby

[root@node1mysql]# crm status

Lastupdated: Wed Jul 9 14:45:36 2014

Lastchange: Wed Jul 9 14:45:29 2014 via crm_attribute on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

3Resources configure

Nodenode1.huhu.com: standby

Online:[ node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node2.huhu.com ]

Stopped:[ node1.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node2.huhu.com

[root@node1mysql]# crm node online

[root@node1mysql]# crm status

Lastupdated: Wed Jul 9 14:45:52 2014

Lastchange: Wed Jul 9 14:45:49 2014 via crm_attribute on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

3Resources configure

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node2.huhu.com ]

Slaves:[ node1.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node2.huhu.com

[root@node1mysql]#

# scp /root/mysql-5.5.38-linux2.6-x86_64.tar.gz node2:/root/

# scp /etc/my.cnf node2:/etc/my.cnf

# scp /etc/init.d/mysqld node2:/etc/init.d/mysqld

在node2上安裝MySQL

#tar zxvf mysql-5.5.38-linux2.6-x86_64.tar.gz -C /usr/local/

#cd /usr/local/

#ln -s mysql-5.5.38-linux2.6-x86_64/ mysql

#cd mysql

#chown root:mysql -R .

注意：千萬不要手動創(chuàng)建/mydata/data，否則會導(dǎo)致文件損壞

缺少相關(guān)庫文件# yum install libaio

#service mysqld start

#/usr/local/mysql/bin/mysql -uroot -e "SHOW DATABASES"

+--------------------+

|Database |

+--------------------+

|information_schema |

|mydb |

|mysql |

|performance_schema |

|test |

+--------------------+

[root@node2mydata]#

#service mysqld stop

#chkconfig mysqld off

配置MySQL成為集群資源

crm(live)#configure

crm(live)configure#primitive mysqld lsb:mysqld

crm(live)configure#verify

crm(live)configure# colocation mysqld_with_mystore inf: mysqld mystore

crm(live)configure#show xml

<rsc_colocation id="mysqld_with_mystore"score="INFINITY" rsc="mysqld" with-rsc="mystore"/>

MySQL服務(wù)一定是跟MySQL存儲資源在一起

crm(live)configure#order mysqld_after_mystore mandatory: mystore mysqld

crm(live)configure#verify

MySQL服務(wù)一定是在MySQ存儲轉(zhuǎn)移之后的，因此定義順序約束

crm(live)#status

Lastupdated: Wed Jul 9 16:18:27 2014

Lastchange: Wed Jul 9 16:18:16 2014 via cibadmin on node2.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

4Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node2.huhu.com ]

Slaves:[ node1.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node2.huhu.com

mysqld (lsb:mysqld): Started node2.huhu.com

crm(live)#

因此登錄node2節(jié)點(diǎn)

#/usr/local/mysql/bin/mysql -uroot -e "SHOW DATABASES"

+--------------------+

|Database |

+--------------------+

|information_schema |

| mydb |

|mysql |

|performance_schema |

|test |

+--------------------+

#/usr/local/mysql/bin/mysql -uroot -e "DROP DATABASE mydb"

#/usr/local/mysql/bin/mysql -uroot -e "CREATE DATABASE testdb"

這里再次進(jìn)行主從節(jié)切換

#crm node standby

#crm status

Masters:[ node1.huhu.com ]

Stopped:[ node2.huhu.com ]

mystore(ocf::heartbeat:Filesystem): Started node1.huhu.com

mysqld(lsb:mysqld): Started node1.huhu.com

#crm node online

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node1.huhu.com ]

Slaves:[ node2.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node1.huhu.com

mysqld (lsb:mysqld): Started node1.huhu.com

在node1節(jié)點(diǎn)上

#/usr/local/mysql/bin/mysql -uroot -e "SHOW DATABASES"

+--------------------+

|Database |

+--------------------+

|information_schema |

|mysql |

|performance_schema |

|test |

|testdb |

+--------------------

testdb正常顯示

***給MySQL定義一個(gè)虛擬IP資源

crm(live)configure#primitive myip ocf:heartbeat:IPaddr paramsip=172.16.100.119 nic=eth0 cidr_netmask=24

crm(live)configure#verify

crm(live)configure#colocation myip_with_MS_mysql_drbd inf:MS_mysql_drbd:Master myip

crm(live)configure#verify

crm(live)configure#show xml

crm(live)configure#commit

crm(live)configure#cd ..

crm(live)#status

Lastupdated: Wed Jul 9 16:46:27 2014

Lastchange: Wed Jul 9 16:46:20 2014 via cibadmin on node1.huhu.com

Stack:classic openais (with plugin)

CurrentDC: node2.huhu.com - partition with quorum

Version:1.1.10-14.el6_5.3-368c726

2Nodes configured, 2 expected votes

5Resources configured

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node1.huhu.com ]

Slaves:[ node2.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node1.huhu.com

mysqld (lsb:mysqld): Started node1.huhu.com

myip (ocf::heartbeat:IPaddr): Started node1.huhu.com

crm(live)

可以看到myip已經(jīng)在node1啟動。

#ip addr

1:lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN

link/loopback00:00:00:00:00:00 brd 00:00:00:00:00:00

inet127.0.0.1/8 scope host lo

inet6::1/128 scope host

valid_lftforever preferred_lft forever

2:eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast stateUP qlen 1000

link/ether00:0c:29:a9:86:42 brd ff:ff:ff:ff:ff:ff

inet172.16.100.103/24 brd 172.16.100.255 scope global eth0

inet172.16.100.119/24 brd 172.16.100.255 scopeglobal secondary eth0

inet6fe80::20c:29ff:fea9:8642/64 scope link

valid_lftforever preferred_lft forever

5.在其節(jié)點(diǎn)上進(jìn)行MySQL登錄驗(yàn)證

登錄MySQL創(chuàng)建用戶

#/usr/local/mysql/bin/mysql -uroot -e "GRANT ALL ON *.* TO root@'%'IDENTIFIED BY '123.com';FLUSH PRIVILEGES"

#mysql -uroot -p123.com -h272.16.100.119 -e "SHOW DATABASES"

+--------------------+

|Database |

+--------------------+

|information_schema |

|mysql |

|performance_schema |

|test |

| testdb |

+--------------------+

[root@localhost~]#

在模擬主從節(jié)點(diǎn)切換：

#crm node standby

#crm node online

#crm status

Online:[ node1.huhu.com node2.huhu.com ]

Master/SlaveSet: MS_mysql_drbd [mysql_drbd]

Masters: [ node2.huhu.com ]

Slaves:[ node1.huhu.com ]

mystore (ocf::heartbeat:Filesystem): Started node2.huhu.com

mysqld (lsb:mysqld): Started node2.huhu.com

myip (ocf::heartbeat:IPaddr): Started node2.huhu.com

#mysql -uroot -p123.com -h272.16.100.119 -e "SHOW DATABASES"

[root@node2~]# crm

crm(live)# configure

crm(live)configure# show

nodenode1.huhu.com \

       attributes standby=off

nodenode2.huhu.com \

       attributes standby=off

primitive myipIPaddr \

       params ip=172.16.100.119 nic=eth0 cidr_netmask=24

primitive mysql_drbdocf:linbit:drbd \

       params drbd_resource=mydrbd \

       op start timeout=240 interval=0 \

       op stop timeout=100 interval=0 \

       op monitor role=Masterinterval=50s timeout=30s \

       op monitor role=Slaveinterval=60s timeout=30s

primitive mysqldlsb:mysqld

primitive mystoreFilesystem \

       params device="/dev/drbd0" directory="/mydata" fstype=ext3 \

       op start timeout=60 interval=0 \

       op stop timeout=60 interval=0

msMS_mysql_drbd mysql_drbd \

       meta master-max=1 master-node-max=1 clone-max=2 clone-node-max=1 notify=true

colocationmyip_with_MS_mysql_drbd inf: MS_mysql_drbd:Master myip

colocationmysqld_with_mystore inf: mysqld mystore

colocationmystore_with_MS_mysql_drbd inf: mystore MS_mysql_drbd:Master

ordermysqld_after_mystore Mandatory: mystore mysqld

ordermystore_after_MS_mysql_drbd Mandatory: MS_mysql_drbd:promote mystore:start

property cib-bootstrap-options: \

       dc-version=1.1.10-14.el6_5.3-368c726\

       cluster-infrastructure="classic openais (with plugin)" \

       expected-quorum-votes=2 \

       stonith-enabled=false\

       no-quorum-policy=ignore

rsc_defaults rsc-options: \

       resource-stickiness=100

crm(live)configure#

以上是“CoroSync+Drbd+MySQL如何實(shí)現(xiàn)MySQL的高可用集群”這篇文章的所有內(nèi)容，感謝各位的閱讀！希望分享的內(nèi)容對大家有幫助，更多相關(guān)知識，歡迎關(guān)注億速云行業(yè)資訊頻道！