lsz_qh 发表于 2014-3-11 10:47:34

exadata x2-2 half rack cell01更换一颗cpu后无法启动

本帖最后由 lsz_qh 于 2014-3-11 10:55 编辑

EXADATA X2-2 HALF RACK

ilom检查发现cell01有一颗cpu报错,exackhk也提示需要更换cpu,换完cpu后,cell01启动报错,把坏的那颗CPU又换回去,系统能正常启动。附件是启动时的系统日志和启动时的屏幕截图。


ilom错误信息:
Are you sure you want to start /SP/faultmgmt/shell (y/n)? y

faultmgmtsp> fmadm faulty
------------------- ------------------------------------ -------------- --------
Time                UUID                                 msgid          Severity
------------------- ------------------------------------ -------------- --------
2013-12-18/19:54:27 8ba2cf4d-8189-e7dc-aeb9-9e4af615aeb8 SPX86-8000-F4  Critical

Fault class : fault.cpu.intel.internal

FRU         : /SYS/MB/P0
              (Part Number: 060C)
              (Serial Number: unknown)

Description : An internal fault on a processor has occurred.

Response    : If running Solaris, the processor will be off-lined upon
              system reboot to prevent future interruptions.

Impact      : System will panic and reset and system performance may be
              impacted.

Action      : The administrator should review the ILOM event log for
              additional information pertaining to this diagnosis.  Please
              refer to the Details section of the Knowledge Article for
              additional information.

Maclean Liu(刘相兵 发表于 2014-3-11 15:36:29

是仅仅更换了硬件吗?
重启前打过 XD的补丁吗?

lsz_qh 发表于 2014-3-11 15:43:07

没有打过补丁,只是更换了一块报错的CPU。当时有2个cell报cpu错误,2个cell都换了,一个cell换完以后能正常启动,cell01就无法启动。

Maclean Liu(刘相兵 发表于 2014-3-11 15:43:11

ODM FINDING:

I think you are referring to the "SR 3-6534026381  post BP11 patch we see Failed to load libsysfs.so.2.0.2"

I have asked the customer to follow the below action plan and this has fixed the issuer

Please see action undertaken below:-



These are the notes to undertake after the bp11 installation on the cells and compute nodes.



WORKAROUND FIX AS PER BUG# 12413272 ) >



#  mount -o loop /u01/PATCHES/11.2.3.2_ISO_YUM_Repo/112320_base_repo.iso /mnt/iso/yum/unknown/EXADATA/dbserver/11.2.3.2.0/base



# ls /mnt/iso/yum/unknown/EXADATA/dbserver/11.2.3.2.0/base/x86_64/L*

/mnt/iso/yum/unknown/EXADATA/dbserver/11.2.3.2.0/base/x86_64/Lib_Utils-1.00-09.noarch.rpm



# cd /mnt/iso/yum/unknown/EXADATA/dbserver/11.2.3.2.0/base/x86_64/



# rpm -Uvh --force --nodeps  Lib_Utils-1.00-09.noarch.rpm

Preparing...                ###########################################

Installing....

   1:Lib_Utils              ###########################################

#



After doing the above =>



# rpm --verify Lib_Utils

# echo $?

0



# /opt/MegaRAID/MegaCli/MegaCli64 -h | head -4

MegaCLI SAS RAID Management Tool  Ver 8.02.21 Oct 21, 2011



# ls -ltr /opt/lsi/3rdpartylibs/

total 140

-rwxr-xr-x 1 root root 97948 Oct 14  2010 libsysfs.so.2.0.2

-r--r--r-- 1 root root 31058 Oct 14  2010 LGPLLicenseV2.txt

drwxr-xr-x 2 root root  4096 Dec 11 14:28 x86_64

drwxr-xr-x 2 root root  4096 Dec 11 14:28 src

lrwxrwxrwx 1 root root    39 Dec 11 14:28 libsysfs.so.2 -> /opt/lsi/3rdpartylibs/libsysfs.so.2.0.2

lrwxrwxrwx 1 root root    35 Dec 11 14:28 libsysfs.so -> /opt/lsi/3rdpartylibs/libsysfs.so.2





# find / -name libsysfs.so.2.0.2

/opt/lsi/3rdpartylibs/x86_64/libsysfs.so.2.0.2

/opt/lsi/3rdpartylibs/libsysfs.so.2.0.2
页: [1]
查看完整版本: exadata x2-2 half rack cell01更换一颗cpu后无法启动