高可用集群之heartbeat配置，实现web服务的高可用以及Mariadb高可用
实验配置环境：# uname -ro
2.6.32-358.el6.i686 GNU/Linux
1.修改节点名称，集群的每个节点名称够能够相互解析
1.修改节点node1主机名
# node1.magedu.com
2.修改节点node1主机名
# node2.magedu.com
3.在集群节点node1和node2的/etc/hosts添加如下两行，使其能够相互解析
# vim /etc/hosts
172.16.0.4 node1.magedu.com node1
172.16.0.7 node2.magedu.com node2
2.配置ntp服务器
1.查看是否安装ntp服务的软件包
# rpm -qa ntp
ntp-4.2.4p8-3.el6.centos.i686
2.将原来配置文件备份，提供新的编辑配置文件/etc/ntp.conf，
# vim /etc/ntp.conf
driftfile /var/lib/ntp/drift
restrict default kod nomodify notrap nopeer noquery
restrict -6 default kod nomodify notrap nopeer noquery
restrict 127.0.0.1
restrict -6 ::1
restrict 172.16.0.0 netmask 255.255.0.0 nomodify notrap
servercn.pool.ntp.orgprefer
server133.100.11.8
server127.127.1.0 # local clock
fudge 127.127.1.0 stratum 10
includefile /etc/ntp/crypto/pw
keys /etc/ntp/keys
logfile /var/log/ntpstats/ntp.log
3.启动nfp服务
# service ntpd start
4.验证ntp服务
# ntpstat
synchronised to local net at stratum 11
time correct to within 11 ms
polling server every 1024 s
# ntpstat
synchronised to local net at stratum 11
time correct to within 11 ms
polling server every 1024 s
# ntpq -p
remote refid st t when poll reach delay offsetjitter
==============================================================================
dns1.synet.edu. 202.118.1.46 2 u134 1024377 95.42814611.6 2042.24
133.100.11.8 133.100.53.51 2 u306 1024377151.02514205.3 2064.42
*LOCAL(0) .LOCL. 10 l 30 64377 0.000 0.000 0.001
3.保证节点一和节点的时间同步
在节点node1和节点node2上配置周期性任务计划，使其时间保持同步
# crontab -e
*/3 * * * * /usr/sbin/ntpdate 172.16.0.6 &> /dev/null
4.各节点之间基于ssh密钥认证通信
节点node1和节点node2不需要密码进行通信
1.将节点node1生成的公钥传到节点node2
# ssh-keygen -t rsa -P ''
# ssh-copy-id -i .ssh/id_rsa.pub root@node1
2.将节点node2生成的公钥传到节点node1
# ssh-keygen -t rsa -P ''
# ssh-copy-id -i .ssh/id_rsa.pub root@node2
3.节点node1和node2时间同步
# ssh node1 'date';date
Sun Apr 20 20:32:00 CST 2014
Sun Apr 20 20:32:00 CST 2014
5.安装配置heartbeat v2
1.解决依赖关系（版本不同解决依赖关系所安装的包有所不同）
# yuminstall PyXML -y
# yuminstall libnet -y
# yum install net-snmp-libs -y
# rpm -ivh heartbeat-2.1.4-12.el6.i686.rpm
# rpm -ivh heartbeat-pils-2.1.4-12.el6.i686.rpm
# rpm -ivh heartbeat-stonith-2.1.4-12.el6.i686.rpm
2.复制heartbeat集群服需要的文件
# cp /usr/share/doc/heartbeat-2.1.4/{haresources,ha.cf,authkeys} /etc/ha.d/
# ls
authkeysharc rc.d resource.d
ha.cf haresourcesREADME.configshellfuncs
3.配置authkeys
# openssl rand -hex 8
cc7cb7ada0d820fb
添加如下两行
# vim authkeys
auth 2
2 sha1 cc7cb7ada0d820fb
修改authkeys的权限，必须为600或400否则不能启动
# chmod 600 authkeys
3.配置核心配置文件ha.cf
logfile /var/log/ha-log 启用日志文件
#logfacility local0
keepalive 800ms 发送心跳信息的间隔
deadtime 4 集群故障之后节点退出集群的时间间隔
warntime 3 警告时间
initdead 90 集群服务启动时的死亡时间间隔
udpport 694 监听端口
mcast eth0 225.0.19.1 694 1 0 启用多播（确保网卡支持多波)
node node1.magedu.com
node node2.magedu.com添加集群节点只能使用主机名
auto_failback on 启用集群节点故障自动转回
ping 172.16.0.1 启用ping节点
compression bz2 启用压缩功能
compression_threshold 2 超过指定的值时则压缩传输，单位KB
4.定义集群资源，使用httpd测试
4.1在node1和node2节点中安装httpd
# yum install httpd -y
# ssh node1 'yum install httpd -y'
在node1和node2确保httpd开启不能自启
# chkconfig httpd off
# chkconfig --list httpd
httpd 0:off 1:off 2:off 3:off 4:off 5:off 6:off
# ssh node1 'chkconfig httpd off'
# ssh node1 'chkconfig --list httpd'
httpd 0:off 1:off 2:off 3:off 4:off 5:off 6:off
4.2分别为node1和node2通过测试页
# vim /var/www/html/index.html
Cluster Node2
# vim /var/www/html/index.html
Cluter Node1
4.3 在node1和node2中启动httpd服务，并测试，然后手动停止服务
#ssh node1 ' service httpd start'; service httpd start
# ssh node1 'curl node2';curl node1
% Total % Received % XferdAverage Speed Time Time TimeCurrent
DloadUpload Total Spent LeftSpeed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0 Cluster Node2
0 24 0 24 0 0 302 0 --:--:-- --:--:-- --:--:-- 400
Cluter Node1
# ssh node1 'service httpd stop';service httpd stop;
4.4 # vim haresources 定义httpd资源
node1.magedu.com 172.16.0.19/16/eth1 httpd
5.将haresources,ha.cf,authkeys，原封不动的复制到node1中
# scp -p authkeys ha.cf haresources node1:/etc/ha.d/
# ssh node1 'ls -l /etc/ha.d/'
total 48
-rw-------. 1 root root 657 Apr 20 22:00 authkeys
-rw-r--r--. 1 root root 10591 Apr 20 22:00 ha.cf
-rwxr-xr-x. 1 root root 745 Sep 102013 harc
-rw-r--r--. 1 root root5963 Apr 20 22:00 haresources
drwxr-xr-x. 2 root root4096 Apr 20 20:48 rc.d
-rw-r--r--. 1 root root 692 Sep 102013 README.config
drwxr-xr-x. 2 root root4096 Apr 20 20:48 resource.d
-rw-r--r--. 1 root root7862 Sep 102013 shellfuncs
6.启动node1和node2的集群服务
# ssh node1 'service heartbeat start';service heartbeat start
# ss -untl | grep 80
tcp LISTEN 0 128 :::80 :::*
7.验证集群服务：
http://s3.运维网.com/wyfs02/M00/24/CB/wKioL1NVEEOx0cEyAADj1p9jv6E226.jpg
关闭node1中的web服务，web服务正常启动，且vip自动转到inode1
http://s3.运维网.com/wyfs02/M00/24/CB/wKioL1NVEhDBypSGAAEDF8e490E174.jpg
http://s3.运维网.com/wyfs02/M01/24/CB/wKioL1NVEqjDtpcYAAJmBvmoOs0638.jpg
8.基于heartbeat-gui配置Mariadb高可用集群服务
所有heartbeat的配置和上面的相同
1.安装heartbeat-gui
rpm -ivh heartbeat-gui-2.1.4-12.el6.i686.rpm
2.在node1和node2的ha.cf添加如下一行
# vim /etc/ha.d/ha/cf
crm on
3.在node1和node2节点中安装Mariadb（具体安装和配置可以参考以前的博客）
4.在node1和node2确保httpd开启不能自启
# chkconfig mysqld off
# chkconfig --list mysqld
mysqld 0:off 1:off 2:off 3:off 4:off 5:off 6:off
# ssh node2 'chkconfig mysqld off'
# ssh node2 'chkconfig --list mysqld'
mysqld 0:off 1:off 2:off 3:off 4:off 5:off 6:off
4.关闭node1和node2中的heartbeat，保证mysqld服务关闭
# ssh node2 'service heartbeat stop';service heartbeat stop
# ssh node2 'service mysqld status';service mysqld status
MySQL is not running
MySQL is not running
5.重启node1和node2中的heartbeat服务，确保保证httpd服务关闭
# ssh node2 'service heartbeat start';service heartbeat start
# ssh node2 'service mysqld status';service mysqld status
MySQL is not running
MySQL is not running
6.验证是否启用heartbeat-gui
# ss -untl | grep 5560
tcp LISTEN 0 10 *:5560 *:*
7.启用heartbeat-gui配置
# hb_gui &
9. 配置Mariadb高可用
http://s3.运维网.com/wyfs02/M02/24/EF/wKiom1NXItWDGSNgAACZ6ZzBjlk049.jpg
http://s3.运维网.com/wyfs02/M00/24/F2/wKioL1NXKeSSe8d_AAFcponyg70005.jpg
10.添加资源
添加虚拟ip首先在Resources选项中选择添加本地资源
http://s3.运维网.com/wyfs02/M02/24/FC/wKiom1NX4z_yO3NGAAC6xd361tA679.jpg
http://s3.运维网.com/wyfs02/M02/24/FC/wKioL1NX5Mry6-jUAAIHuBJleLo970.jpg
http://s3.运维网.com/wyfs02/M02/24/FC/wKiom1NX5pXxe1MTAAF_kB7pq-0972.jpg
添加mysqld，直接点击add就能够添加Mariadb
http://s3.运维网.com/wyfs02/M01/24/FC/wKioL1NX56TD_sQPAAEK_s0p2OY086.jpg
定义资源约束
添加位置约束使vip和Mariadb更倾向node1节点
http://s3.运维网.com/wyfs02/M00/24/FC/wKiom1NX6dmAaZlBAAFiKA0cXK0557.jpg
http://s3.运维网.com/wyfs02/M02/24/FC/wKioL1NX65Xwu9grAAJce4c4NjA033.jpg
定义顺序约束，指定资源启动和关闭的顺序
http://s3.运维网.com/wyfs02/M00/24/FC/wKioL1NX7MyjIvciAADduTHgzQU077.jpg
定义Mariadb和vip必须运行于同一个节点
http://s3.运维网.com/wyfs02/M01/24/FC/wKioL1NX7xaAQ9wfAAEHM9mf6Is369.jpg
11.启动Mariadb高可用服务
http://s3.运维网.com/wyfs02/M00/24/FC/wKiom1NX8ZmhrhpXAAETSqjQi04710.jpg
12.验证配置的Mariadb高可用服务，使node1离线并且能够实现故障转移
http://s3.运维网.com/wyfs02/M02/24/FC/wKioL1NX8SKTR2l3AACNdIyNKto726.jpg
http://s3.运维网.com/wyfs02/M01/24/FC/wKiom1NX8iOwnBrZAAIL0mfhK5Y844.jpg
总结：
heartbeat v1和v2必须在每一个节点都必须在每一节点上都进行相同的配置，配
置高可用主要是实现服务的高可用，一般用几个9来衡量。配置高可用集群必须修改每
个节点的名称，使其能够解析，并且每个集群节点的时间必须保持同步。当出现如下错
误时：使用hostname命令修改为响应的主机即可。
service heartbeat start
Starting High-Availability services:
Heartbeat failure . Failed.
heartbeat: udpport setting must precede media statementsheartbeat: 2014/04/23_23:04:40 info: Version 2 support: on
heartbeat: 2014/04/23_23:04:40 WARN: File /etc/ha.d/haresources exists.
heartbeat: 2014/04/23_23:04:40 WARN: This file is not used because crm is enabled
heartbeat: baudrate setting must precede media statementsheartbeat: 2014/04/23_23:04:40 ERROR: Current node not in configuration!
heartbeat: 2014/04/23_23:04:40 info: By default, cluster nodes are named by `uname -n` and must be declared with a 'node' directive in the ha.cf file.
heartbeat: 2014/04/23_23:04:40 info: See also: http://linux-ha.org/ha.cf/NodeDirective
heartbeat: 2014/04/23_23:04:40 WARN: Logging daemon is disabled --enabling logging daemon is recommended
heartbeat: 2014/04/23_23:04:40 ERROR: Configuration error, heartbeat not started.
页:
[1]