pacemaker,是一个群集资源管理器。它实现最大可用性群集服务(亦称资源管理)的节点和资源级故障检测和恢复使用您的首选集群基础设施(OpenAIS的或Heaerbeat)提供的消息和成员能力。
它可以做乎任何规模的集群,并配备了一个强大的依赖模型,使管理员能够准确地表达群集资源之间的关系(包括顺序和位置)。几乎任何可以编写脚本,可以管理作为心脏起搏器集群的一部分。
pacemaker的特点:
主机和应用程序级别的故障检测和恢复
几乎支持任何冗余配置
同时支持多种集群配置模式
配置策略处理法定人数损失(多台机器失败时)
支持应用启动/关机顺序
支持,必须/必须在同一台机器上运行的应用程序
支持多种模式的应用程序(如主/从)
可以测试任何故障或群集的群集状态
primitive(native):基本资源,原始资源
group:资源组
clone:克隆资源(可同时运行在多个节点上),要先定义为primitive后才能进行clone。主要包含STONITH和集群文件系统(cluster filesystem)
master/slave:主从资源,如drdb
Lsb:linux表中库,一般位于/etc/rc.d/init.d/目录下的支持start|stop|status等参数的服务脚本都是lsb
ocf:Open cluster Framework,开放集群架构
heartbeat:heartbaet V1版本
stonith:专为配置stonith设备而用
172.25.85.2 server2.example.com(1024)
172.25.85.3 server3.example.com(1024)
1. server2/3:
/etc/init.d/ldirectord stop
chkconfig ldirectord off
yum install pacemaker -y
rpm -q corosync
server2:
cd /etc/corosync
cp corosync.conf.example corosync.conf
vim /etc/corosync/corosync.conf
bindnetaddr: 172.25.85.0 service { name: pacemaker ver: 0 }
scp /etc/corosync/corosync.conf root@172.25.85.3:/etc/corosync/
/etc/init.d/corosync start
crm_verify -LV ##检测配置文件是否正确
server3:
/etc/init.d/corosync start
tail -f /var/log/messages
2. server2:
yum install crmsh-1.2.6-0.rc2.2.1.x86_64.rpm pssh-2.3.1-2.1.x86_64.rpm
crm
crm(live)configure# property no-quorum-policy=ignore ##将修改同步到其他节点 crm(live)configure# commit
vim /etc/httpd/conf/httpd.conf
<Location /server-status> SetHandler server-status Order deny,allow Deny from all Allow from 127.0.0.1 </Location>
crm
crm(live)configure# primitive vip ocf:heartbeat:IPaddr2 params ip=172.25.85.100 crm(live)configure# primitive vip ocf:heartbeat:IPaddr2 params ip=172.25.85.100 cidr_netmask=32 op monitor interval=30s primitive website ocf:heartbeat:apache params configfile=/etc/httpd/conf/httpd.conf op monitor interval=60s crm(live)configure# commit
crm(live)configure# collocation website-with-ip inf: website vip ##将这两个资源整合到一个主机上
crm(live)configure# commit
crm(live)configure# delete website-with-ip
crm(live)configure# commit
crm(live)configure# group apache vip website ##创建一个资源组apache
crm(live)configure# commit
crm(live)node# show server2.example.com: normal server3.example.com: normal crm(live)node# standby server2.example.com
crm(live)resource# stop apache crm(live)resource# show Resource Group: apache vip (ocf::heartbeat:IPaddr2): Stopped website (ocf::heartbeat:apache): Stopped crm(live)configure# delete apache crm(live)configure# delete website crm(live)configure# commit
[注意]
删除一个资源:
crm(live)resource# stop website
crm(live)configure# delete website
server3:
yum install crmsh-1.2.6-0.rc2.2.1.x86_64.rpm pssh-2.3.1-2.1.x86_64.rpm
vim /etc/httpd/conf/httpd.conf
<Location /server-status> SetHandler server-status Order deny,allow Deny from all Allow from 127.0.0.1 </Location>
crm_mon ##
3.真机上:systemctl start fence_virtd.service
server2:
stonith_admin -I
stonith_admin -M -a fence_xvm
crm
crm(live)configure# property stonith-enabled=true
primitive vmfence stonith:fence_xvm params pcmk_host_map="server2.example.com:ricci1;server3.example.com:heartbeat1" op monitor interval=1min
primitive sqldata ocf:linbit:drbd params drbd_resource=sqldata op monitor interval=30s
ms sqldataclone sqldata meta master-max=1 master-node-max=1 clone-max=2 clone-node-max=1 notify=true
crm(live)configure# commit
crm_mon
primitive sqlfs ocf:heartbeat:Filesystem params device=/dev/drbd1 directory=/var/lib/mysql fstype=ext4
colocation sqlfs_on_drbd inf: sqlfs sqldataclone:Master
order sqlfs-after-sqldata inf: sqldataclone:promote sqlfs:start
crm(live)configure# commit
crm_mon
crm(live)configure# primitive mysql lsb:mysqld op monitor interval=30s crm(live)configure# group mygroup vip sqlfs mysql INFO: resource references in colocation:sqlfs_on_drbd updated INFO: resource references in order:sqlfs-after-sqldata updated
crm(live)configure# commit
crm_mon
crm(live)resource# refresh mygroup
server3:
crm_mon ##监控节点状态
原文地址:http://11713145.blog.51cto.com/11703145/1834577