/usr/sbin/bmc-watchdog[2451]: fiid_obj_get:

系统不定时重启，检查日志报以下warning：
/usr/sbin/bmc-watchdog[2451]: fiid_obj_get: 'timer_state': data not available

原因：package “freeipmi-bmc-watchdog”引起

 

 

解决办法参考：
Install RHEL 6.5 with package freeipmi-bmc-watchdog.  Boot into the operating system and then the system will Auto Restart periodically.  The /var/log/messages log file will display the following message if the issue occurs:

/usr/sbin/bmc-watchdog[XXXX]: fiid_obj_get: 'present_countdown_value': data not available

NOTE:  In the BMC log, there will be records saying “watchdog hard reset”.

Affected configurations

The above symptom is associated with, but not limited to, the following systems:

• ThinkServer TS440
• ThinkServer RD340/RD440/RD540/RD640/RD550/RD650
• ThinkServer TD350

System is configured with:

• None

Operating Systems:

• RedHat Enterprise Linux Server 6.5

Limitation

According to the RedHat errata article below, this is a known issue which can be fixed by updating freepmi packages.

When the Intelligent Platform Management Interface (IPMI) service was not configured and kernel modules were not loaded and the bmc-watchdog timer was started, the freeipmi bmc-watchdog could cause an unexpected system reboot on certain hardware.

The new update adds the module alias configuration file for use with the module-init-tools package.  This configuration provides an early kernel module probing rule for systems where the IPMI is advertised through an Advanced Configuration and Power Interface (ACPI) entry.  These rules help to avoid an occasional reset with IPMI service misconfiguration.

Solution

Solution 1:

• Do not install the package “freeipmi-bmc-watchdog”.

Solution 2:

• Install the package “freeipmi-bmc-watchdog”.
• Disable service bmc-watchdog by manually selecting “chkconfig bmc-watchdog off” and then reboot.

Solution 3:

• Install the RedHat newest package “freeipmi-1.2.1-6” which will resolve this issue.

转载至http://blog.chinaunix.net/uid-26313110-id-5163796.html