Chinaunix

标题: 求助HP rx6600升级微码报错 [打印本页]

作者: tomorrowlm 时间: 2013-06-20 11:36
标题: 求助HP rx6600升级微码报错
HP rx6600 升级微码后报警，报警灯亮，可以正常启动操作系统，以下是报警日志：

Log Entry 36: 19 Jun 2013 23:19:03
Alert Level 7: Fatal
Keyword: MC_INITIATED
Machine Check initiated
Logged by: System Firmware  6
Data: Major change in system state - HPMC or MCA
0xF480009806E00320 000000000000000B


Log Entry 35: 19 Jun 2013 23:19:03
Alert Level 7: Fatal
Keyword: MACHINE_CHECK_INITIATED
Machine Check initiated
Logged by: Redundant w/ an E0 code;
Sensor: Critical Interrupt
Data2: OEM Code2: 0x00
0xC151C23C67020310 003FA17000130300

作者: haizdl 时间: 2013-06-20 19:04
〉〉Keyword: MACHINE_CHECK_INITIATED

发生MCA了，收集MCA日志，找惠普工具Decode。。。。。。。。

作者: lbseraph 时间: 2013-06-20 21:34
回复 1# tomorrowlm

MCA一般是硬件问题导致的。进EFI shell里面抓取MCA日志，或看下/var/tombstones目录下有没有相应的mca文件，收取到相应日志后找HP分析一下。
shell> errdump mca

PS: 一般你系统都起来了，而又没有再次出现类似MCA的话，基本没什么问题，可以继续观察。

作者: NEDK 时间: 2013-06-26 21:17
看看是不是每次都报MC吧，不一定有问题，如果每次都报的话，肯定是硬件问题

作者: lbseraph 时间: 2013-06-28 23:35
嗯，还有如果正常能进到系统后消掉告警灯后再观察吧，如果真是硬件问题导致，那么还会出现MCA的报错，那时候就可以抓日志和MCA来看了。

作者: cjhvslhb 时间: 2013-06-30 18:17
IPMI Event Code: F480009806E00320 000000000000000B

Record Type       = E0h
Reporting Entity ID = System Firmware - cpu 6
Event ID          = #152

...........................................................

Keyword          = MC_INITIATED

Description:

A Machine Check has been initiated

Cause / Action:

A Machine Check has occurred.

Recommendation:

Analyze cause of Machine Check using diag's and EFI tools.

___________________________________________________________

Alert Level       = 7  - Fatal
Data Type          = 20 - Major change in state
                        For details, see 4.2.20 of Event Architecture Specification
Data             = 00 00 00 00 00 00 00 0B
_______________________________________________________________________________

IPMI Event Code: C151C23C67020310 003FA17000130300

Timestamp (GMT) = Wed Jun 19 23:19:03 2013
Generator       = System Software ID = 0
Alert Level    = 7 - Fatal

Sensor
Number    Triplet    Data 2    Data 3
------ --------    ------    ------
00    13:70:A1       3F       00

Decoding as system type: Ruby/Sapphire

Sensor Number : 00 - Various, refer to manual

Sensor Type : 13 = Critical Interrupt

Event type = Assertion event : OEM defined
Keyword = MACHINE_CHECK_INITIATED
Machine Check initiated.  Processor = 00
Go to the EFI shell and do

Shell> errdump mca 'filename'

where filename is any name you like.  This file can then be run
through the MCA analyser tool

A Machine Check Abort event means the hardware detected a critical error. This event is generated
whenever a system error due to processor, firmware, hardware and operating system is encountered.
MCA events may be either recoverable or non-recoverable. If it is recoverable, the system will
attempt to recover from the error for the purpose of maintaining high availability. An example of
which is automatic disabling of a failing processor. For non-recoverable errors, the system will
either stop or reboot to prevent data corruption and unreliable operation

你把MCA的文件和你的升级过程日志发过来看看呢

欢迎光临 Chinaunix (http://bbs.chinaunix.net/)