Operating System - OpenVMS
cancel
Showing results for 
Search instead for 
Did you mean: 

MACHINE CHECK IN KERNEL MODE

mrityunjoy
Advisor

MACHINE CHECK IN KERNEL MODE

Hi,

 

Yesterday on server (ES40 , openvms 7.3-2) was rebooted with Machine check. I have attached the error log and cue dump. Could you please guide me why it happened?

 

 

 

Mrityunjoy Kundu -AST (TCS)
6 REPLIES
Volker Halle
Honored Contributor

Re: MACHINE CHECK IN KERNEL MODE

Mrityunjoy,

 

you need to extract the MACHINECHK errlog entries from the dump.

 

$ **bleep**/CRASH dumpfile

SDA> CLUE ERRLOG

SDA> EXIT

 

Then try $ DIAGNOSE CLUE$ERRLOG.SYS. Note that DECevent does NOT support ES40, so you might need WEBES/SEA to decode a ES40 machine check errlog entry.

 

Please post both the output of SDA> CLUE ERRLOG and the DECevent analysis of CLUE$ERRLOG.SYS. A CLUE file is not sufficient to determine the reason for a MACHINECHK crash, you NEED the ERRLOG entries immediately preceeding the crash.

 

Volker.

mrityunjoy
Advisor

Re: MACHINE CHECK IN KERNEL MODE

Hi Volker,

 

Please find the required information.

 

Dumpfile Errorlog Entry Information:
------------------------------------
Sequence  Date        Time            Error Message Type
--------  ----------- -----------     --------------------------------
    3530   9-AUG-2012 18:33:09.74     Machine Check 670
    3531   9-AUG-2012 18:33:09.74  *  Crash Entry

Mrityunjoy Kundu -AST (TCS)
Volker Halle
Honored Contributor

Re: MACHINE CHECK IN KERNEL MODE

Mrityunjoy,

 

as I expected: DECevent V3.3 did not decode the ES40 Machine Check 670 error from CLUE$ERRLOG.SYS. You need WEBES/SEA to decode an ES40 machine check error.

 

Can you try to attach the CLUE$ERRLOG.SYS file (after zipping it) ? Then I can run that file through SEA and provide the results.

 

Volker.

mrityunjoy
Advisor

Re: MACHINE CHECK IN KERNEL MODE

Hi Volker,

 

Sorry for late reply. Please find the sys file attached.

Mrityunjoy Kundu -AST (TCS)
Volker Halle
Honored Contributor

Re: MACHINE CHECK IN KERNEL MODE

Mrityunjoy,

 

here is the WEBES/SEA analysis (V4.4.4 Build 18).

 

Volker.

 

Brief Description:

CPU0 Bcache tag parity event detected.

 

Full Description:

     An uncorrectable CPU0 Bcache tag parity infraction has been diagnosed. 
     ***  While this event was fatal, it is not recommended that a FRU
     replacement be done on the first occurrence of the failure. There is a
     very low probability of a reoccurrence on the same hardware component. 
     FRU information has been provided to allow failure correlation in the
     unlikely event of a repeat failure. Please discuss this failure with your
     support center for the correct course of action.

FRU List:

     Probability:        High
     Manufacturer:       HP
     Device Type:        Central Processing Unit
     Physical Location:  Slot CPU 0 -- J40.
     FRU Part Number:    54-30362-B3.C01   
     FRU Serial Number:  AY14001561
     FRU Firmware Rev:   00

     Probability:        Low
     Manufacturer:       HP
     Device Type:        System Main Unit
     Physical Location:  Slot Central System Motherboard
     FRU Part Number:    54-25385-02.A01   
     FRU Serial Number:  AY13310263
     FRU Firmware Rev:   TIG 10|SROM V2.22-G|SRM V7.3-1|RMC V10|RMC FLASH V27
mrityunjoy
Advisor

Re: MACHINE CHECK IN KERNEL MODE

Thanks  Volker.

Mrityunjoy Kundu -AST (TCS)