服务器：HP DL385 G7

操作系统：suse10 sp3

数据库：oracle 11g R2

集群软件：VCS 双机主备

环境：两台服务器使用VCS软件做的oracle主备切换数据库

故障现象：

1.两台数据库主机不定期频繁重启，每次重启时在操作系统message日志中均没有任何记录；

2.系统启动时，message 日志出现与硬件相关的错误信息

 message 日志信息：

-------------------------------------------------------------------------------------------------------------

Oct 27 17:51:01 linux10 /usr/sbin/cron[5968]: (CRON) STARTUP (V5.0)

Oct 27 17:51:02 linux10 sshd[6047]: Server listening on 0.0.0.0 port 22.

Oct 27 17:51:02 linux10 rcpowersaved: CPU frequency scaling is not supported by your processor.

Oct 27 17:51:02 linux10 rcpowersaved: enter 'CPUFREQ_ENABLED=no' in /etc/powersave/cpufreq to avoid this warning.

Oct 27 17:51:02 linux10 rcpowersaved: Cannot load cpufreq governors - No cpufreq driver available

Oct 27 17:51:03 linux10 rcpowersaved: s2ram does not know your machine. See 's2ram -i' for details. (127)

Oct 27 17:51:03 linux10 rcpowersaved: Use SUSPEND2RAM_FORCE=yes to override this detection.

Oct 27 17:51:03 linux10 modprobe: FATAL: Error running install command for binfmt_misc

Oct 27 17:51:06 linux10 kernel: klogd 1.4.1, log source = /proc/kmsg started.

Oct 27 17:51:06 linux10 kernel: Floppy drive(s): fd0 is 1.44M

Oct 27 17:51:06 linux10 syslog-ng[5762]: Changing permissions on special file /dev/xconsole

Oct 27 17:51:06 linux10 syslog-ng[5762]: Changing permissions on special file /dev/tty10

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-10 - disabling barriers

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-11 - disabling barriers

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-12 - disabling barriers

Oct 27 17:51:06 linux10 kernel: AppArmor: AppArmor initialized

Oct 27 17:51:06 linux10 kernel: audit(1414403451.182:2):  info="AppArmor initialized" pid=4403

Oct 27 17:51:06 linux10 kernel: floppy0: no floppy controllers found

Oct 27 17:51:06 linux10 kernel: ACPI: Power Button (FF) [PWRF]

Oct 27 17:51:06 linux10 kernel: rdac: device handler unregistered

Oct 27 17:51:06 linux10 kernel: No dock devices found.

Oct 27 17:51:06 linux10 kernel: bnx2: eth0: using MSI

Oct 27 17:51:06 linux10 kernel: bnx2: eth1: using MSI

Oct 27 17:51:06 linux10 kernel: Ethernet Channel Bonding Driver: v3.2.5 (March 21, 2008)

Oct 27 17:51:06 linux10 kernel: bonding: Warning: either miimon or arp_interval and arp_ip_target module parameters must be s

pecified, otherwise bonding will not detect link failures! see bonding.txt for details.

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-8 - disabling barriers

Oct 27 17:51:06 linux10 kernel: bnx2: eth0 NIC Copper Link is Up, 1000 Mbps full duplex

Oct 27 17:51:06 linux10 kernel: bonding: bond0: setting mode to active-backup (1).

Oct 27 17:51:06 linux10 kernel: bonding: bond0: Setting MII monitoring interval to 100.

Oct 27 17:51:06 linux10 kernel: bonding: bond0: Setting use_carrier to 0.

Oct 27 17:51:06 linux10 kernel: bnx2: eth0: using MSI

Oct 27 17:51:06 linux10 kernel: bonding: bond0: enslaving eth0 as a backup interface with a down link.

Oct 27 17:51:06 linux10 kernel: bnx2: eth1: using MSI

Oct 27 17:51:06 linux10 kernel: bonding: bond0: enslaving eth1 as a backup interface with a down link.

Oct 27 17:51:06 linux10 kernel: audit(1414403461.814:3): audit_pid=5906 old=0 by auid=4294967295

Oct 27 17:51:06 linux10 kernel: llt: module not supported by Novell, setting U taint flag.

Oct 27 17:51:06 linux10 kernel: LLT INFO V-14-1-10009 LLT 5.1.100.000-SP1GA Protocol available

Oct 27 17:51:06 linux10 kernel: bnx2: eth1 NIC Copper Link is Up, 1000 Mbps full duplex

Oct 27 17:51:06 linux10 kernel: bonding: bond0: link status definitely up for interface eth1.

Oct 27 17:51:06 linux10 kernel: bonding: bond0: making interface eth1 the new active one.

Oct 27 17:51:06 linux10 kernel: bonding: bond0: first active interface up!

Oct 27 17:51:06 linux10 kernel: bnx2: eth0 NIC Copper Link is Up, 1000 Mbps full duplex

Oct 27 17:51:06 linux10 kernel: powernow-k8: Found 4 AMD Opteron(tm) Processor 6134 processors (16 cpu cores) (version 2.20.0

0)

Oct 27 17:51:06 linux10 kernel: powernow-k8: MP systems not supported by PSB BIOS structure

…………

Oct 28 18:10:01 linux10 /usr/sbin/cron[17099]: (root) CMD (/usr/sbin/ntpdate 172.29.141.162)

Oct 28 18:11:14 linux10 zmd: ShutdownManager (WARN): Preparing to sleep...

Oct 28 18:11:15 linux10 zmd: ShutdownManager (WARN): Going to sleep, waking up at 10/29/2014 17:41:08

Oct 28 18:11:49 linux10 syslog-ng[5762]: Error connecting to remote host AF_INET(172.29.141.162:5140), reattempting in 60 sec

onds

…………

-----------------------------------------------------------------------

在上面的日志中出现两个问题分别为：

一、zmd: ShutdownManager (WARN): Preparing to sleep…

二、

Oct 27 17:51:02 linux10 rcpowersaved: CPU frequency scaling is not supported by your processor.

Oct 27 17:51:02 linux10 rcpowersaved: enter 'CPUFREQ_ENABLED=no' in /etc/powersave/cpufreq to avoid this warning.

Oct 27 17:51:02 linux10 rcpowersaved: Cannot load cpufreq governors - No cpufreq driver available

…………

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-10 - disabling barriers

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-11 - disabling barriers

Oct 27 17:51:06 linux10 kernel: JBD: barrier-based sync failed on dm-12 - disabling barriers

 

问题一 由ZMD服务器引起，Novell对ZMD服务的解释为：

The zmd daemon performs software management functions on the ZENworks managed device, including updating, installing, and removing software, and performing basic queries of the device's package management database. Typically, these management tasks are initiated through the ZENworks Control Center or the rug, zen-installer, zen-updater, or zen-remover utilities, which means you should not need to interact directly with zmd.

ZMD服务主要负责用户软件的更新、安装管理操作，在开机时自动启动，ZMD服务启动后，默认每六小时联网更新，更新时会占用80端口，因此经常会与tomcat 等服务器产生端口，因此在软件安装或更新完后，可以及时关闭此服务，

#/etc/init.d/novell-zmd status

Checking for ZENworks Management Daemon:                              running

#/etc/init.d/novell-zmd stop

Shutting down ZENworks Management Daemon                              done

注：关闭 此服务后，安装软件是比较麻烦，因此在需要时可以在此打开，改服务在更新时有可能会长时间锁定/etc/mtab，因此需要注意。

 解决方法：

 关闭novell-zmd服务后，此日志消失。

有时我们为了提高开机速度，会将novell-zmd服务进行关闭

chkconfig -delete novell-zmd

 

问题二：

单从日志信息上看cpu不支持变频的问题，由于在操作系统和VCS日志中均没有发现其他异常，因此怀疑是服务器硬件出了问题，去机房一看，服务器住面板有电流符号的故障灯显示橘红色，这时基本就能放松了，硬件肯定是不对了，于是收集硬件日志联系HP厂商，经确定是主板故障，更换主板后，服务器没有重启。