CentOS 6.5上部署Heartbeat-阿里云开发者社区

环境说明：

主机名

角色

IP地址

VIP

heartbeat01.contoso.com

Heartbeat节点1

eth0:192.168.49.133

eth1:172.16.49.133(心跳连接)

172.16.49.100

heartbeat02.contoso.com

Heartbeat节点2

eth0:192.168.49.134

eth1:172.16.49.134(心跳连接)

一、准备工作

以下操作除非特别指明，否则均需在两台服务器上操作。

 
        # 关闭iptables防火墙并禁用SELinux
       
        /etc/init
        .d
        /iptables 
        stop 
       
        chkconfig iptables off
       
        sed 
        -i 
        '/^SELINUX/s/enforcing/disabled/' 
        /etc/selinux/config 
       
        setenforce 0
       
        # 设置时间同步
       
        crontab 
        -e  
        # 添加计划任务 
       
        0 * * * * 
        /usr/sbin/ntpdate   
        210.72.145.44 64.147.116.229 
        time
        .nist.gov 
       
        :wq
       
        或者
       
        echo 
        '0 * * * * /usr/sbin/ntpdate   210.72.145.44 64.147.116.229 time.nist.gov' 
        >>
        /var/spool/cron/root  
        # 添加计划任务 
       
        crontab 
        -l  
        # 检查计划任务是否存在 
       
        0 * * * * 
        /usr/sbin/ntpdate   
        210.72.145.44 64.147.116.229 
        time
        .nist.gov 
       
        # 设置主机名（以heartbeat01为例，heartbeat02同样的方法）
       
        sed 
        -i 
        '/^HOSTNAME/s/^/#/' 
        /etc/sysconfig/network  
       
        sed 
        -i 
        '/#HOSTNAME/aHOSTNAME=heartbeat01.contoso.com' 
        /etc/sysconfig/network 
       
        grep 
        HOSTNAME 
        /etc/sysconfig/network 
       
        hostname 
        heartbeat01.contoso.com 
       
        或者
       
        sed 
        -i 
        '/^HOSTNAME/d' 
        /etc/sysconfig/network 
       
        echo 
        'HOSTNAME=heartbeat01.contoso.com' 
        >>
        /etc/sysconfig/network 
       
        grep 
        HOSTNAME 
        /etc/sysconfig/network 
       
        hostname 
        heartbeat01.contoso.com 
       
        # 编辑/etc/hosts文件
       
        echo 
        -e 
        '192.168.49.133  heartbeat01.contoso.com\n192.168.49.134  heartbeat02.contoso.com' 
        >>
        /etc/hosts 
       
        tail 
        -2 
        /etc/hosts 
       
        # 添加一条主机路由
       
        /sbin/route 
        add -host 172.16.49.134 dev eth1 
        # 在heartbeat01上配置 
       
        echo 
        '/sbin/route add -host 172.16.49.134 dev eth1' 
        >>
        /etc/rc
        .
        local 
        # 在heartbeat01上配置 
       
        /sbin/route 
        add -host 172.16.49.133 dev eth1 
        # 在heartbeat02上配置 
       
        echo 
        '/sbin/route add -host 172.16.49.133 dev eth1' 
        >>
        /etc/rc
        .
        local 
        # 在heartbeat02上配置 
       
        route -n 
        #添加之后分别在heartbeat01和heartbeat02上检查

二、安装heartbeat软件

 
        rpm -ivh http:
        //dl
        .fedoraproject.org
        /pub/epel/6/x86_64/epel-release-6-8
        .noarch.rpm 
       
        yum -y 
        install 
        heartbeat*

三、编辑heartbeat配置文件

1）拷贝配置文件

 
        cp 
        /usr/share/doc/heartbeat-3
        .0.4/{ha.cf,haresources,authkeys} 
        /etc/ha
        .d/ 
       
        ll 
        /etc/ha
        .d/ 
       
        cd 
        /etc/ha
        .d/

2）配置authkeys

[root@heartbeat01 ha.d]# egrep -v "#|^$" authkeys

auth 2

2 sha1 c6091592594cd14c

[root@heartbeat02 ha.d]# egrep -v "#|^$" authkeys

auth 2

2 sha1 c6091592594cd14c

# 两个节点的配置一致

3）配置ha.cf

下面以使用单播的方式为例，给出两个节点的配置：

[root@heartbeat01 ha.d]# egrep -v "#|^$" ha.cf

debugfile /var/log/ha-debug #设置debug文件位置

logfile /var/log/ha-log #设置日志文件位置

logfacility local1 #设置记录日志的设备

keepalive 2 #设置发送心跳报文的时间间隔

deadtime 30 #设置确认对端死亡的时间间隔

warntime 10 #设置发出最后的心跳警告报文的间隔

initdead 60 #设置初始化时间

ucast eth1 172.16.49.134 #设定侦听的心跳线的接口和对应的对端接口的IP地址

auto_failback on #启用自动恢复模式，当拥有该资源的属主恢复之后，属主将回收该资源

node heartbeat01.contoso.com #指定节点1，节点的名称一定要和uname -n的结果一致

node heartbeat02.contoso.com #指定节点2

ping 172.16.49.1 #指定第三方仲裁节点

respawn hacluster /usr/lib64/heartbeat/ipfail #使用这个脚本去侦听对方是否还活着（使用的是ICMP报文检测）

[root@heartbeat02 ha.d]# egrep -v "#|^$" ha.cf

debugfile /var/log/ha-debug

logfile /var/log/ha-log

logfacility local1

keepalive 2

deadtime 30

warntime 10

initdead 60

ucast eth1 172.16.49.133

auto_failback on

node heartbeat01.contoso.com

node heartbeat02.contoso.com

ping 172.16.49.1

respawn hacluster /usr/lib64/heartbeat/ipfail

# 两个节点的差别只有单播的对端IP不一样，其他都一样

4）配置haresources

echo 'heartbeat01.contoso.com IPaddr::172.16.49.100/24/eth1' >>/etc/ha.d/haresources

[root@heartbeat01 ha.d]# egrep -v "#|^$" haresources

heartbeat01.contoso.com IPaddr::172.16.49.100/24/eth1

[root@heartbeat02 ha.d]# egrep -v "#|^$" haresources

heartbeat01.contoso.com IPaddr::172.16.49.100/24/eth1

# 两个节点的配置一致

四、启动heartbeat并测试

 
        /etc/init
        .d
        /heartbeat 
        start  
        #分别在heartbeat01和heartbeat02上执行

到两个节点上分别查看VIP：

[root@heartbeat01 ha.d]# ip addr |grep 172.16.49

inet 172.16.49.133/24 brd 172.16.49.255 scope global eth1

inet 172.16.49.100/24 brd 172.16.49.255 scope global secondary eth1

[root@heartbeat02 ha.d]# ip addr |grep 172.16.49

inet 172.16.49.134/24 brd 172.16.49.255 scope global eth1

然后，将heartbeat01上的heartbeat服务关闭，再进行查看：

[root@heartbeat01 ha.d]# /etc/init.d/heartbeat stop

Stopping High-Availability services: Done.

[root@heartbeat01 ha.d]# ip addr |grep 172.16.49

inet 172.16.49.133/24 brd 172.16.49.255 scope global eth1

[root@heartbeat02 ha.d]# ip addr |grep 172.16.49

inet 172.16.49.134/24 brd 172.16.49.255 scope global eth1

inet 172.16.49.100/24 brd 172.16.49.255 scope global secondary eth1

可以看到，VIP已经从heartbeat01上转移到heartbeat02上了。

在VIP切换过程中，从另一台主机ping VIP地址，间断时间非常短暂。

五、检查日志

/var/log/ha-log

================================================

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Comm_now_up(): updating status to active

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Local status now set to: 'active'

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Starting child client "/usr/lib64/heartbeat/ipfail" (498,499)

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7312]: info: Starting "/usr/lib64/heartbeat/ipfail" as uid 498 gid 499 (pid 7312)

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Status update for node heartbeat02.contoso.com: status active

harc(default)[7315]: 2016/09/22_05:44:26 info: Running /etc/ha.d//rc.d/status status

Sep 22 05:44:33 heartbeat01.contoso.com ipfail: [7312]: info: Asking other side for ping node count.

Sep 22 05:44:36 heartbeat01.contoso.com ipfail: [7312]: info: No giveup timer to abort.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7284]: info: remote resource transition completed.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7284]: info: Initial resource acquisition complete (T_RESOURCES(us))

/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.49.100)[7368]: 2016/09/22_05:44:37 INFO: Resource is stopped

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7332]: info: Local Resource acquisition completed.

harc(default)[7451]: 2016/09/22_05:44:37 info: Running /etc/ha.d//rc.d/ip-request-resp ip-request-resp

ip-request-resp(default)[7451]: 2016/09/22_05:44:37 received ip-request-resp IPaddr::172.16.49.100/24/eth1 OK yes

ResourceManager(default)[7474]: 2016/09/22_05:44:37 info: Acquiring resource group: heartbeat01.contoso.com IPaddr::172.16.49.100/24/eth1

/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.49.100)[7502]: 2016/09/22_05:44:37 INFO: Resource is stopped

ResourceManager(default)[7474]: 2016/09/22_05:44:37 info: Running /etc/ha.d/resource.d/IPaddr 172.16.49.100/24/eth1 start

IPaddr(IPaddr_172.16.49.100)[7627]: 2016/09/22_05:44:38 INFO: Adding inet address 172.16.49.100/24 with broadcast address 172.16.49.255 to device eth1

IPaddr(IPaddr_172.16.49.100)[7627]: 2016/09/22_05:44:38 INFO: Bringing device eth1 up

IPaddr(IPaddr_172.16.49.100)[7627]: 2016/09/22_05:44:38 INFO: /usr/libexec/heartbeat/send_arp -i 200 -r 5 -p /var/run/resource-agents/send_arp-172.16.49.100 eth1 172.16.49.100 auto not_used not_used

/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.49.100)[7601]: 2016/09/22_05:44:38 INFO: Success

Sep 22 05:44:40 heartbeat01.contoso.com heartbeat: [7284]: info: Heartbeat shutdown in progress. (7284)

Sep 22 05:44:40 heartbeat01.contoso.com heartbeat: [7716]: info: Giving up all HA resources.

ResourceManager(default)[7729]: 2016/09/22_05:44:40 info: Releasing resource group: heartbeat01.contoso.com IPaddr::172.16.49.100/24/eth1

ResourceManager(default)[7729]: 2016/09/22_05:44:40 info: Running /etc/ha.d/resource.d/IPaddr 172.16.49.100/24/eth1 stop

IPaddr(IPaddr_172.16.49.100)[7792]: 2016/09/22_05:44:40 INFO: IP status = ok, IP_CIP=

/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.49.100)[7766]: 2016/09/22_05:44:40 INFO: Success

Sep 22 05:44:40 heartbeat01.contoso.com heartbeat: [7716]: info: All HA resources relinquished.

Sep 22 05:44:41 heartbeat01.contoso.com heartbeat: [7284]: WARN: 1 lost packet(s) for [heartbeat02.contoso.com] [20:22]

Sep 22 05:44:41 heartbeat01.contoso.com heartbeat: [7284]: info: No pkts missing from heartbeat02.contoso.com!

Sep 22 05:44:41 heartbeat01.contoso.com heartbeat: [7284]: info: killing /usr/lib64/heartbeat/ipfail process group 7312 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: killing HBFIFO process 7288 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: killing HBWRITE process 7289 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: killing HBREAD process 7290 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: killing HBWRITE process 7291 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: killing HBREAD process 7292 with signal 15

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: Core process 7292 exited. 5 remaining

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: Core process 7289 exited. 4 remaining

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: Core process 7290 exited. 3 remaining

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: Core process 7291 exited. 2 remaining

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: Core process 7288 exited. 1 remaining

Sep 22 05:44:43 heartbeat01.contoso.com heartbeat: [7284]: info: heartbeat01.contoso.com Heartbeat shutdown complete.

/var/log/ha-debug

================================================

Sep 22 05:44:14 heartbeat01.contoso.com heartbeat: [7283]: info: **************************

Sep 22 05:44:14 heartbeat01.contoso.com heartbeat: [7283]: info: Configuration validated. Starting heartbeat 3.0.4

Sep 22 05:44:14 heartbeat01.contoso.com heartbeat: [7284]: info: heartbeat: version 3.0.4

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: Heartbeat generation: 1474533038

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: write socket priority set to IPTOS_LOWDELAY on eth1

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: bound send socket to device: eth1

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: set SO_REUSEPORT(w)

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: bound receive socket to device: eth1

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: set SO_REUSEPORT(w)

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ucast: started on port 694 interface eth1 to 172.16.49.134

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: glib: ping heartbeat started.

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: G_main_add_TriggerHandler: Added signal manual handler

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: G_main_add_SignalHandler: Added signal handler for signal 17

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: Local status now set to: 'up'

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: Link 172.16.49.1:172.16.49.1 up.

Sep 22 05:44:15 heartbeat01.contoso.com heartbeat: [7284]: info: Status update for node 172.16.49.1: status ping

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Link heartbeat02.contoso.com:eth1 up.

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Status update for node heartbeat02.contoso.com: status up

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7294]: debug: notify_world: setting SIGCHLD Handler to SIG_DFL

harc(default)[7294]: 2016/09/22_05:44:26 info: Running /etc/ha.d//rc.d/status status

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Comm_now_up(): updating status to active

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Local status now set to: 'active'

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Starting child client "/usr/lib64/heartbeat/ipfail" (498,499)

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: debug: get_delnodelist: delnodelist=

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7312]: info: Starting "/usr/lib64/heartbeat/ipfail" as uid 498 gid 499 (pid 7312)

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7284]: info: Status update for node heartbeat02.contoso.com: status active

Sep 22 05:44:26 heartbeat01.contoso.com heartbeat: [7315]: debug: notify_world: setting SIGCHLD Handler to SIG_DFL

harc(default)[7315]: 2016/09/22_05:44:26 info: Running /etc/ha.d//rc.d/status status

Sep 22 05:44:26 heartbeat01.contoso.com ipfail: [7312]: debug: PID=7312

Sep 22 05:44:26 heartbeat01.contoso.com ipfail: [7312]: debug: Signing in with heartbeat

Sep 22 05:44:27 heartbeat01.contoso.com ipfail: [7312]: debug: [We are heartbeat01.contoso.com]

Sep 22 05:44:27 heartbeat01.contoso.com ipfail: [7312]: debug: auto_failback -> 1 (on)

Sep 22 05:44:27 heartbeat01.contoso.com ipfail: [7312]: debug: Setting message filter mode

Sep 22 05:44:28 heartbeat01.contoso.com ipfail: [7312]: debug: Starting node walk

Sep 22 05:44:29 heartbeat01.contoso.com ipfail: [7312]: debug: Cluster node: 172.16.49.1: status: ping

Sep 22 05:44:29 heartbeat01.contoso.com ipfail: [7312]: debug: Cluster node: heartbeat02.contoso.com: status: active

Sep 22 05:44:30 heartbeat01.contoso.com ipfail: [7312]: debug: [They are heartbeat02.contoso.com]

Sep 22 05:44:30 heartbeat01.contoso.com ipfail: [7312]: debug: Cluster node: heartbeat01.contoso.com: status: active

Sep 22 05:44:31 heartbeat01.contoso.com ipfail: [7312]: debug: Setting message signal

Sep 22 05:44:31 heartbeat01.contoso.com ipfail: [7312]: debug: Waiting for messages...

Sep 22 05:44:32 heartbeat01.contoso.com ipfail: [7312]: debug: Got join message from another ipfail client. (heartbeat02.contoso.com)

Sep 22 05:44:33 heartbeat01.contoso.com ipfail: [7312]: debug: Found ping node 172.16.49.1!

Sep 22 05:44:33 heartbeat01.contoso.com ipfail: [7312]: info: Asking other side for ping node count.

Sep 22 05:44:33 heartbeat01.contoso.com ipfail: [7312]: debug: Message [num_ping] sent.

Sep 22 05:44:36 heartbeat01.contoso.com ipfail: [7312]: info: No giveup timer to abort.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7284]: info: remote resource transition completed.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7284]: info: Initial resource acquisition complete (T_RESOURCES(us))

Sep 22 05:44:37 heartbeat01.contoso.com ipfail: [7312]: debug: Other side is now stable.

/usr/lib/ocf/resource.d//heartbeat/IPaddr(IPaddr_172.16.49.100)[7368]: 2016/09/22_05:44:37 INFO: Resource is stopped

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7332]: info: Local Resource acquisition completed.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7284]: debug: StartNextRemoteRscReq(): child count 1

Sep 22 05:44:37 heartbeat01.contoso.com ipfail: [7312]: debug: Other side is now stable.

Sep 22 05:44:37 heartbeat01.contoso.com heartbeat: [7451]: debug: notify_world: setting SIGCHLD Handler to SIG_DFL