ipmi fence问题请教大家

winson1740

内核：2.6.18-194.el5
OS：Red Hat Enterprise Linux Server release 5.5 (Tikanga)
2台IBM x3650 m2做RHCS，用x3650 m2服务器的ipmi做fence设备
网络设置：
节点1：
public IP 10.72.86.121
private IP 10.1.1.1
ipmi IP 10.72.86.126


节点2：
public IP 10.72.86.122
private IP 10.1.1.2
ipmi IP 10.72.86.127

心跳private IP网口用直连网线连接起来，public IP网口和ipmiIP网口接到同一个交换机

现象1：在节点1上用fence_ipmilan 10.72.86.127命令可以fence节点2,让节点2重启了:
[root@elndb1 /]# fence_ipmilan -a 10.72.86.127
Rebooting machine @ IPMI:10.72.86.127...Done
但是messages日志会报如下错误信息：
Jun 28 19:47:17 elndb1 fenced[6977]: agent "fence_ipmilan" reports: Rebooting machine @ IPMI:10.72.86.127...Failed
Jun 28 19:47:17 elndb1 fenced[6977]: fence "elndb2.eln.com" failed


现象2：在节点1上用fence_node elndb2.eln.com命令fence节点2失败：
[root@elndb1 ~]# fence_node elndb2.eln.com
agent "fence_ipmilan" reports: Rebooting machine @ IPMI:10.72.86.127...Failed
messages日志报如下错误信息：
Jun 28 20:43:11 elndb1 fence_node[8830]: agent "fence_ipmilan" reports: Rebooting machine @ IPMI:10.72.86.127...Failed
Jun 28 20:43:11 elndb1 fence_node[8830]: Fence of "elndb2.eln.com" was unsuccessful



cluster.conf文件配置如下：
<?xml version="1.0"?>
<cluster alias="elndb_cluster" config_version="9" name="elndb_cluster">
<fence_daemon post_fail_delay="0" post_join_delay="3"/>
<clusternodes>
<clusternode name="elndb1.eln.com" nodeid="1" votes="1">
<fence>
<method name="1">
<device name="ipmi1"/>
</method>
</fence>
</clusternode>
<clusternode name="elndb2.eln.com" nodeid="2" votes="1">
<fence>
<method name="1">
<device name="ipmi2"/>
</method>
</fence>
</clusternode>
</clusternodes>
<cman expected_votes="1" two_node="1"/>
<fencedevices>
<fencedevice agent="fence_ipmilan" auth="none" ipaddr="10.72.86.126" login="USERID" name="ipmi1" passwd="PASSW0RD"/>
<fencedevice agent="fence_ipmilan" auth="none" ipaddr="10.72.86.127" login="USERID" name="ipmi2" passwd="PASSW0RD"/>
</fencedevices>
<rm>
<failoverdomains>
<failoverdomain name="elndbdomain" restricted="1">
<failoverdomainnode name="elndb1.eln.com" priority="1"/>
<failoverdomainnode name="elndb2.eln.com" priority="1"/>
</failoverdomain>
</failoverdomains>
<resources>
<ip address="10.72.86.130" monitor_link="1"/>
</resources>
<service autostart="0" domain="elndbdomain" name="elndb_svc" recovery="relocate">
<ip ref="10.72.86.130"/>
</service>
</rm>
</cluster>


从messages日志看到的信息非常有限，请问一下大家该如何处理？

xiaobing927

该问题属于你的USERID的权限不够，所以fence失败。

xiaobing927

还有一个可能是：
<fencedevice agent="fence_ipmilan" auth="none" ipaddr="10.72.86.127" login="USERID" name="ipmi2" passwd="PASSW0RD"/>

中的auth="none"应该删掉，存在这个会有问题。另外：你的用户和密码是系统默认的特权用户应该是可以的