Heartbeat--为什么起动不了haresources里面的资源?-<已解决>

Question

软件:heartbeat-2.0.5.tar.gz---------安装过程没有问题
我的haresources的主要内容:
Loadbalance IPaddr:192.168.0.20/32/192.168.0.20 lvs mon

说明
Loadbalance----主director的主机名字
192.168.0.20---virtual IP
lvs---/etc/init.d/lvs---一个启动IPVS规则的脚本文件
mon--/etc/init.d/mon---监视realserver的脚本

特殊说明mon一定要这样运行才生效:

1. #/etc/init.d/mon &

复制代码
而不能

1. #/etc/init.d/mon start

复制代码
本人水平非常菜,不能使mon以#/etc/init.d/mon start的方式运行,郁闷一

lvs可以这样运行

1. #/etc/init.d/lvs start

复制代码

问题:
当我
#/etc/init.d/heartbeat start的时候

1. root@Loadbalance:/etc/ha.d# /etc/init.d/heartbeat start
2. Starting High-Availability services:
3. Done.

复制代码

但是lvs和mon都没有运行起来,不知道是什么原因?

[ 本帖最后由 Linuxcn.com 于 2006-6-27 10:05 编辑 ]

Answer 1

原帖由 zhenh 于 2006-6-23 22:52 发表
ha.cf 里把 ucast eth0 192.168.0.11 换成 bcast eth0
haresources 变成 “Loadbalance 192.168.0.20"
你的 uname -n 的结果是 "loadbalance" 还是 "Loadbalance"? 大小写要分清。

...

把haresources 变成 “Loadbalance 192.168.0.20",运行的结果是:

1. root@Loadbalance:/etc/ha.d# /etc/init.d/heartbeat start
2. Starting High-Availability services:
3. 2006/06/25_05:56:05 INFO: IPaddr Resource is stopped
4. Done.

复制代码

虚拟IP没起来

而我单独这样

1. root@Loadbalance1:/etc/ha.d/resource.d# ./IPaddr 192.168.0.20 start
2. 2006/06/25_07:15:13 INFO: /sbin/ifconfig eth0:0 192.168.0.20 netmask 255.255.255.0 broadcast 192.168.0.255
3. 2006/06/25_07:15:13 INFO: Sending Gratuitous Arp for 192.168.0.20 on eth0:0 [eth0]
4. 2006/06/25_07:15:13 INFO: /usr/lib/heartbeat/send_arp -i 500 -r 10 -p /var/run/heartbeat/rsctmp/send_arp/send_arp-192.168.0.20 eth0 192.168.0.20 auto 192.168.0.20 ffffffffffff
5. 2006/06/25_07:15:13 INFO: IPaddr Success
6. INFO: IPaddr Success

复制代码

再来看log,Loadbalance的log:

1. heartbeat[19586]: 2006/06/25_06:19:32 info: Configuration validated. Starting heartbeat 2.0.5
2. heartbeat[19587]: 2006/06/25_06:19:32 info: heartbeat: version 2.0.5
3. heartbeat[19587]: 2006/06/25_06:19:32 info: Heartbeat generation: 5
4. heartbeat[19587]: 2006/06/25_06:19:32 info: G_main_add_TriggerHandler: Added signal manual handler
5. heartbeat[19587]: 2006/06/25_06:19:32 info: G_main_add_TriggerHandler: Added signal manual handler
6. heartbeat[19587]: 2006/06/25_06:19:32 info: Removing /var/run/heartbeat/rsctmp failed, recreating.
7. heartbeat[19587]: 2006/06/25_06:19:32 info: glib: UDP Broadcast heartbeat started on port 694 (694) interface eth0
8. heartbeat[19587]: 2006/06/25_06:19:32 info: glib: UDP Broadcast heartbeat closed on port 694 interface eth0 - Status: 1
9. heartbeat[19587]: 2006/06/25_06:19:32 info: G_main_add_SignalHandler: Added signal handler for signal 17
10. heartbeat[19587]: 2006/06/25_06:19:32 info: Local status now set to: 'up'
11. heartbeat[19587]: 2006/06/25_06:19:33 info: Link loadbalance:eth0 up.
12. heartbeat[19587]: 2006/06/25_06:20:35 info: Link loadbalance1:eth0 up.
13. heartbeat[19587]: 2006/06/25_06:20:35 info: Status update for node loadbalance1: status up
14. harc[19596]:    2006/06/25_06:20:35 info: Running /etc/ha.d/rc.d/status status
15. heartbeat[19587]: 2006/06/25_06:20:36 info: Comm_now_up(): updating status to active
16. heartbeat[19587]: 2006/06/25_06:20:36 info: Local status now set to: 'active'
17. heartbeat[19587]: 2006/06/25_06:20:36 info: Status update for node loadbalance1: status active
18. harc[19607]:    2006/06/25_06:20:36 info: Running /etc/ha.d/rc.d/status status
19. heartbeat[19587]: 2006/06/25_06:20:47 info: remote resource transition completed.
20. heartbeat[19587]: 2006/06/25_06:20:47 info: remote resource transition completed.
21. heartbeat[19587]: 2006/06/25_06:20:47 info: Initial resource acquisition complete (T_RESOURCES(us))
22. IPaddr[19644]:  2006/06/25_06:20:49 INFO: IPaddr Resource is stopped
23. heartbeat[19617]: 2006/06/25_06:20:49 info: Local Resource acquisition completed.
24. harc[19743]:    2006/06/25_06:20:49 info: Running /etc/ha.d/rc.d/ip-request-resp ip-request-resp
25. ip-request-resp[19743]: 2006/06/25_06:20:49 received ip-request-resp 192.168.0.20 OK yes
26. ResourceManager[19758]: 2006/06/25_06:20:49 info: Acquiring resource group: loadbalance 192.168.0.20 lvs moon

复制代码

Loadbalance1的log

1. heartbeat[17249]: 2006/06/25_07:33:23 info: Configuration validated. Starting heartbeat 2.0.5
2. heartbeat[17250]: 2006/06/25_07:33:23 info: heartbeat: version 2.0.5
3. heartbeat[17250]: 2006/06/25_07:33:23 info: Heartbeat generation: 13
4. heartbeat[17250]: 2006/06/25_07:33:23 info: G_main_add_TriggerHandler: Added signal manual handler
5. heartbeat[17250]: 2006/06/25_07:33:23 info: G_main_add_TriggerHandler: Added signal manual handler
6. heartbeat[17250]: 2006/06/25_07:33:23 info: Removing /var/run/heartbeat/rsctmp failed, recreating.
7. heartbeat[17250]: 2006/06/25_07:33:23 info: glib: UDP Broadcast heartbeat started on port 694 (694) interface eth0
8. heartbeat[17250]: 2006/06/25_07:33:23 info: glib: UDP Broadcast heartbeat closed on port 694 interface eth0 - Status: 1
9. heartbeat[17250]: 2006/06/25_07:33:23 info: G_main_add_SignalHandler: Added signal handler for signal 17
10. heartbeat[17250]: 2006/06/25_07:33:24 info: Local status now set to: 'up'
11. heartbeat[17250]: 2006/06/25_07:33:25 info: Link loadbalance1:eth0 up.
12. heartbeat[17250]: 2006/06/25_07:33:25 info: Link loadbalance:eth0 up.
13. heartbeat[17250]: 2006/06/25_07:33:25 info: Status update for node loadbalance: status up
14. heartbeat[17250]: 2006/06/25_07:33:25 info: Status update for node loadbalance: status active
15. heartbeat[17250]: 2006/06/25_07:33:25 info: Comm_now_up(): updating status to active
16. heartbeat[17250]: 2006/06/25_07:33:25 info: Local status now set to: 'active'
17. harc[17258]:    2006/06/25_07:33:25 info: Running /etc/ha.d/rc.d/status status
18. harc[17268]:    2006/06/25_07:33:25 info: Running /etc/ha.d/rc.d/status status

复制代码

看起来应该是正常的啊

这样是可以的,迷茫中

[ 本帖最后由 Linuxcn.com 于 2006-6-26 10:50 编辑 ]

Answer 2

ha.cf 里把 ucast eth0 192.168.0.11 换成 bcast eth0
haresources 变成 “Loadbalance 192.168.0.20"
你的 uname -n 的结果是 "loadbalance" 还是 "Loadbalance"? 大小写要分清。

然后两边都起heartbeat. 看看192.168.0.20有没有起来。

Answer 3

想了半天,还是决定顶一下.

Answer 4

原帖由 q1208c 于 2006-6-23 13:03 发表
目前我的能力可能帮不了你太多.

node name 一定要和 uname -n 一样.
后面那个mon的好象可以自己写一个script 来处理的.

无论如何,还是要感谢你的建议啦

node name是和uname -n一样的,也就是主机名
mon我自己已经"抄"+改动,已经可以以/etc/init.d/mon start的方式启动了,而且监视realserver功能也正常

现在就是如何配双机的问题

Answer 5

目前我的能力可能帮不了你太多.

node name 一定要和 uname -n 一样.
后面那个mon的好象可以自己写一个script 来处理的.

Answer 6

最后,顺便请教一下大家,
像/usr/bin/mon &这样运行的服务,我何如把它停掉?

Answer 7

原帖由 q1208c 于 2006-6-22 22:10 发表
最好能把你的 ha.cf 还 ha-log 帖上来看看. 不然不知道为什么.

友性提示 32 位的 netmask 用来做什么呀?

不得不承认,小弟是纯粹照"本"抄来的,32位的netmask好象是用来做广播的,我把虚拟IP这样配置:

1. ifconfig eth0:0 192.168.0.20 netmask 255.255.255.255 broadcast 192.168.0.20

复制代码

客户机是可以正常通过访问虚拟IP192.168.0.20,得到真实服务器提供的web服务的,而且LVS的功能也正常
详情可看小弟的另一个帖子
http://bbs.chinaunix.net/viewthr ... ghlight=Linuxcn.com

原理是什么,小弟真的不清楚,抱歉

ha-log:

1. heartbeat[6568]: 2006/06/22_09:47:23 WARN: Core dumps could be lost if multiple dumps occur
2. heartbeat[6568]: 2006/06/22_09:47:23 WARN: Consider setting /proc/sys/kernel/core_uses_pid (or equivalent) to 1 for maximum supportability
3. heartbeat[6568]: 2006/06/22_09:47:23 WARN: Logging daemon is disabled --enabling logging daemon is recommended
4. heartbeat[6568]: 2006/06/22_09:47:23 info: **************************
5. heartbeat[6568]: 2006/06/22_09:47:23 info: Configuration validated. Starting heartbeat 2.0.5
6. heartbeat[6569]: 2006/06/22_09:47:23 info: heartbeat: version 2.0.5
7. heartbeat[6569]: 2006/06/22_09:47:23 info: Heartbeat generation: 6
8. heartbeat[6569]: 2006/06/22_09:47:23 info: G_main_add_TriggerHandler: Added signal manual handler
9. heartbeat[6569]: 2006/06/22_09:47:23 info: G_main_add_TriggerHandler: Added signal manual handler
10. heartbeat[6569]: 2006/06/22_09:47:23 info: Removing /var/run/heartbeat/rsctmp failed, recreating.
11. heartbeat[6569]: 2006/06/22_09:47:23 info: glib: ucast: write socket priority set to IPTOS_LOWDELAY on eth0
12. heartbeat[6569]: 2006/06/22_09:47:23 info: glib: ucast: bound send socket to device: eth0
13. heartbeat[6569]: 2006/06/22_09:47:23 info: glib: ucast: bound receive socket to device: eth0
14. heartbeat[6569]: 2006/06/22_09:47:23 info: glib: ucast: started on port 694 interface eth0 to 192.168.0.10
15. heartbeat[6569]: 2006/06/22_09:47:23 info: G_main_add_SignalHandler: Added signal handler for signal 17
16. heartbeat[6569]: 2006/06/22_09:47:23 info: Local status now set to: 'up'
17. heartbeat[6569]: 2006/06/22_09:47:25 info: Link loadbalance:eth0 up.
18. heartbeat[6569]: 2006/06/22_09:47:25 info: Status update for node loadbalance: status active
19. harc[6576]:     2006/06/22_09:47:25 info: Running /etc/ha.d/rc.d/status status
20. heartbeat[6569]: 2006/06/22_09:47:25 info: Comm_now_up(): updating status to active
21. heartbeat[6569]: 2006/06/22_09:47:25 info: Local status now set to: 'active'
22. heartbeat[6569]: 2006/06/22_09:47:26 info: remote resource transition completed.
23. heartbeat[6569]: 2006/06/22_09:47:26 info: remote resource transition completed.
24. heartbeat[6569]: 2006/06/22_09:47:26 info: Local Resource acquisition completed. (none)
25. heartbeat[6569]: 2006/06/22_09:47:26 info: loadbalance wants to go standby [foreign]
26. heartbeat[6569]: 2006/06/22_09:47:27 info: standby: acquire [foreign] resources from loadbalance
27. heartbeat[6587]: 2006/06/22_09:47:27 info: acquire local HA resources (standby).
28. heartbeat[6587]: 2006/06/22_09:47:27 info: local HA resource acquisition completed (standby).
29. heartbeat[6569]: 2006/06/22_09:47:27 info: Standby resource acquisition done [foreign].
30. heartbeat[6569]: 2006/06/22_09:47:27 info: Initial resource acquisition complete (auto_failback)
31. heartbeat[6569]: 2006/06/22_09:47:27 info: remote resource transition completed.

复制代码
[ 本帖最后由 Linuxcn.com 于 2006-6-23 09:46 编辑 ]

Answer 8

Loadbalance的ha.cf:

1. logfile /var/log/ha-log
2. logfacility     local0
3. keepalive 2
4. deadtime 30
5. initdead 120
6. udpport 694
7. ucast eth0 192.168.0.11
8. auto_failback on
9. node    Loadbalance
10. node    Loadbalance1

复制代码
而Loadbalance1的ha.cf就只有这句不同,其他都一样:

1. ucast eth0 192.168.0.10

复制代码

haresource现在都改为:

1. Loadbalance 192.168.0.20 lvs moon

复制代码

其中,lvs,moon2个服务都在/etc/init.d/目录下面
我单独这样运行:

1. /etc/init.d/lvs start

复制代码

1. /etc/init.d/moon start

复制代码
ipvsadm的规则,虚拟IP,mon监视程序都可以正常服务的

请老大们帮我看看是那里出了问题

哦,最后要交代的是,我是在VM上做的实验,很多帖子上说要用心跳信号,但是我一直不是很理解他们这么说:

1. 2台director机器都是双网卡:
2. 192.168.0.10   Loadbalance-----LAN
3. 10.0.0.10         Loadbalance-----心跳信号线
4. 192.168.0.11   Loadbalance1----LAN
5. 10.0.0.11         Loadbalance1----心跳信号线
6. 一个串口线-----也是用来做心跳的.........后面省略若干

复制代码

既然已经有了心跳线,为什么还要1个串口线做心跳线呢
而且在ha.cf里他们把

1. baud   19200
2. serial /dev/ttyS0      # Linux
3. bcast  eth1        # Linux
4. mcast eth1 225.0.0.1 694 1 0

复制代码
都选上了

小弟新手,实在搞的有点糊涂了

[ 本帖最后由 Linuxcn.com 于 2006-6-23 09:32 编辑 ]

Answer 9

最好能把你的 ha.cf 还 ha-log 帖上来看看. 不然不知道为什么.

友性提示 32 位的 netmask 用来做什么呀?