heartbeat v2基于haresources实现HA Web
前言:
该文章花费了笔者大量时间编写,转载请回复征得笔者同意!
实现目标:
使用heartbeat和haresources实现Web集群的高可用
术语解释:
heartbeat是什么:
Heartbeat 项目是 Linux-HA 工程的一个组成部分,它实现了一个高可用集群系统
为什么要使用heartbeat:
在Linux中实现服务的高可用需要搭建HA集群,HA集群需要提供Messaging Layer和Cluster Resource Manager功能的软件
实现CRM的组件:
heartbeat V2版本中有两个组件可以实现CRM的功能,分别是haresources,crm.在此我们使用haresources来实现对于各节点资源的管理
实验环境:
System: CentOS6.7
host:
Primary: node1.anyisalin.com
Standby: node2.anyisalin.com
Network Configure:
node1: 172.16.100.7/24 eth0
node2: 172.16.100.8/24 eth0
float IP: 172.16.100.1/24 #对外提供服务的IP
实验步骤:
1.使用hosts文件实现HA集群中的节点能够解析各节点的主机名
2.使用公钥认证实现主机之间互信通信
3.安装配置heartbeat,httpd
4.测试web服务器是否实现高可用
1.使用hosts文件实现HA集群中的节点能够解析各节点的主机名
[root@node2~]#vim/etc/hosts127.0.0.1localhostlocalhost.localdomainlocalhost4localhost4.localdomain4::1localhostlocalhost.localdomainlocalhost6localhost6.localdomain6172.16.100.7node1.anyisalin.com172.16.100.8node2.anyisalin.com[root@node2~]#scp-p/etc/hostsnode1.anyisalin.com:/etc/
2.使用公钥认证实现主机之间互信通信
[root@node2~]#ssh-keygen-trsa-P''-f~/.ssh/id_rsa [root@node2~]#ssh-copy-id-i~/.ssh/id_rsa.pubnode1.anyisalin.com [root@node1~]#ssh-keygen-trsa-P''-f~/.ssh/id_rsa [root@node1~]#ssh-copy-id-t~/.ssh/id_rsanode2.anyisalin.com
3.安装配置heartbeat,httpd
[root@node1~]#wget-O/etc/yum.repos.d/centos.repohttps://lug.ustc.edu.cn/wiki/_export/code/mirrors/help/centos?codeblock=2##使用USTC镜像源 [root@node1~]#yuminstallepel-rpm-macros.noarch-y [root@node1~]#yuminstsallheartbeathttpd-y [root@node1~]#sshnode2.anyisalin.com"yuminstall-yepel-rpm-macros.noarch" [root@node1~]#sshnode2.anyisalin.com"yuminstall-yheartbeathttpd" [root@node1~]#cp/usr/share/doc/heartbeat-3.0.4/{authkeys,ha.cf,haresources}/etc/ha.d/ [root@node1~]#chmod600/etc/ha.d/authkeys
[root@node1~]#vim/etc/ha.d/authkeys##在最后添加两行auth11md5d41d8cd98f00b204e9800998ecf8427e-##保存退出
[root@node1~]#vim/etc/ha.d/ha.cf##在结尾添加几行bcasteth0logfile/var/log/ha-logkeepalive1nodenode1.anyisalin.com#这里的主机名必须和uname-n显示的一致nodenode2.anyisalin.com#同上##保存退出
[root@node1~]#vim/etc/ha.d/haresources##在最后添加一行node1.anyisalin.comIPaddr::172.16.100.1/24/eth0httpd##结束
[root@node1~]#scp/etc/ha.d/{authkeys,haresources,ha.cf}node2.anyisalin.com:/etc/ha.d/ [root@node1~]#echo"ThisisHAnode1">/var/www/html/index.html#创建网页文件 [root@node1~]#sshnode2.anyisalin.com"echo"ThisisHAnode2">/var/www/html/index.html"#为node2创建网页文件 [root@node1~]#serviceheartbeatstart#启动heartbeat服务 [root@node1~]#sshnode2.anyisalin.com"serviceheartbeatstart"#为node2启动heartbeat服务
[root@node1~]#tail/var/log/ha-log#监控日志文件ResourceManager(default)[8766]:2016/01/02_06:22:00info:Running/etc/ha.d/resource.d/IPaddr172.16.100.1/24/eth0startIPaddr(IPaddr_172.16.100.1)[8916]:2016/01/02_06:22:00INFO:Addinginetaddress172.16.100.1/24withbroadcastaddress172.16.100.255todeviceeth0IPaddr(IPaddr_172.16.100.1)[8916]:2016/01/02_06:22:00INFO:Bringingdeviceeth0upIPaddr(IPaddr_172.16.100.1)[8916]:2016/01/02_06:22:00INFO:/usr/libexec/heartbeat/send_arp-i200-r5-p/var/run/resource-agents/send_arp-172.16.100.1eth0172.16.100.1autonot_usednot_used/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.100.1)[8890]:2016/01/02_06:22:00INFO:SuccessResourceManager(default)[8766]:2016/01/02_06:22:00info:Running/etc/init.d/httpdstart
[root@node1~]#ipa##node1为主节点1:lo:mtu65536qdiscnoqueuestateUNKNOWNlink/loopback00:00:00:00:00:00brd00:00:00:00:00:00inet127.0.0.1/8scopehostloinet6::1/128scopehostvalid_lftforeverpreferred_lftforever2:eth0:mtu1500qdiscpfifo_faststateUPqlen1000link/ether00:0c:29:cf:c0:42brdff:ff:ff:ff:ff:ffinet172.16.100.7/24brd172.16.100.255scopeglobaleth0inet172.16.100.1/24brd172.16.100.255scopeglobalsecondaryeth0inet6fe80::20c:29ff:fecf:c042/64scopelinkvalid_lftforeverpreferred_lftforever
4.测试web服务器是否实现高可用
[root@DataBase~]#curl172.16.100.1#测试Web服务器,页面显示为主节点node1的网页文件ThisisHAnode1[root@node1~]#/usr/share/heartbeat/hb_standby#使用heartbeat内置脚本模拟FailOver(故障转移)[root@node1~]#ipa1:lo:mtu65536qdiscnoqueuestateUNKNOWNlink/loopback00:00:00:00:00:00brd00:00:00:00:00:00inet127.0.0.1/8scopehostloinet6::1/128scopehostvalid_lftforeverpreferred_lftforever2:eth0:mtu1500qdiscpfifo_faststateUPqlen1000link/ether00:0c:29:cf:c0:38brdff:ff:ff:ff:ff:ffinet172.16.100.7/24brd172.16.100.255scopeglobaleth0inet6fe80::20c:29ff:fecf:c038/64scopelinkvalid_lftforeverpreferred_lftforever
[root@node2~]#ipa1:lo:mtu65536qdiscnoqueuestateUNKNOWNlink/loopback00:00:00:00:00:00brd00:00:00:00:00:00inet127.0.0.1/8scopehostloinet6::1/128scopehostvalid_lftforeverpreferred_lftforever2:eth0:mtu1500qdiscpfifo_faststateUPqlen1000link/ether00:0c:29:cf:c0:42brdff:ff:ff:ff:ff:ffinet172.16.100.8/24brd172.16.100.255scopeglobaleth0inet172.16.100.1/24brd172.16.100.255scopeglobalsecondaryeth0inet6fe80::20c:29ff:fecf:c042/64scopelinkvalid_lftforeverpreferred_lftforever
[root@DataBase~]#curl172.16.100.1#重新发起请求,页面显示为从节点node2的网页文件ThisisHAnode2
实验总结:
heartbeat和haresources实现Web集群的高可用,在主节点故障的情况下能够通过CRM将资源转移到备用节点上,但是该实验还有很多未考虑的因素,在此不过多进行解释