MCE Help

[samurai_linux]

Does anyone know how to interpret the following MCE log

HARDWARE ERROR. This is *NOT* a software problem!
Please contact your hardware vendor
CPU 4 0 data cache TSC a4d4c13bd54b26
Data cache ECC error (syndrome 2a)
bit46 = corrected ecc error
bit57 = processor context corrupt
bit61 = error uncorrected
memory/cache error 'data write mem transaction, data transaction, level 1'
STATUS b615400000000145 MCGSTATUS 4


Thanks in advance

[Rubberman]

The driver has detected a hardware error in the cache memory. You need to send the board in for repair or replacement.

[samurai_linux]

The driver has detected a hardware error in the cache memory. You need to send the board in for repair or replacement.

Thanks for your response.

Are you saying the processor/memory board to be replaced?

The system is HP Proliant DL585 G2

Do you know how it the MCE is interpreted?

[Rubberman]

Thanks for your response.

Are you saying the processor/memory board to be replaced?

The system is HP Proliant DL585 G2

Do you know how it the MCE is interpreted?

I thought the MCE error message was quite clear. CPU #4 has an unrecoverable cache memory error. Run the system diagnostics, or download them from HP and run them (you will probably need to create a bootable CD or USB drive with the diagnostic package - instructions on how to do that should be on the HP support site). My guess is that if you run the system board tests in a continuous manner, it will fail after awhile.

[samurai_linux]

I thought the MCE error message was quite clear. CPU #4 has an unrecoverable cache memory error. Run the system diagnostics, or download them from HP and run them (you will probably need to create a bootable CD or USB drive with the diagnostic package - instructions on how to do that should be on the HP support site). My guess is that if you run the system board tests in a continuous manner, it will fail after awhile.

Actually I ran the diagnostics on the box couple of times but it came clean every time.
I am suspecting the problem could be with the CPU 4 itself but not sure. I thought if I get to know how to interpret the MCE in some way, I can get a clue or know what's going on.

[Rubberman]

As I said. It seems perfectly clear to me - interpretation not needed. Does this happen frequently, or only once? If only once, it might have been a transient problem - cosmic rays perhaps? If it recurs, then it is likely a hardware problem. It is a CPU problem since it is a level 1 cache error (next to last line). The last line (STATUS) is probably only meaningful to the CPU manufacturer.