question about machne checksum errors
Hi,
Thanks in advance for any
clarifications.
I was trying to analyze a few machine
checksum errors, and obtained a lot of valuable information in the archive, including a referrence to Dave Jones' tool, parsemce. Thanks !
The servers that we get these occasional
errors on, are mostly dual AMD boxes (MP 1800s, typically) running 2.4.18. Are the decoding rules, present in Intel's IA-32 MCA documentation, applicable
on these servers ? As detailed in Intel's Software Developer's Manual, looks like it is not mandatory for an IA-32 compliant system to follow the decoding rules on all the fields in the machine status
registers (e.g the ones that indicate Uncorrectable ECC errors)..
Appreciate any help regarding this.
-thanks,
Atish
__________________________________________________
Do You Yahoo!?
Yahoo! Finance - Get real-time stock quotes
http://finance.yahoo.com
Quoted 4 lines · Atish Datta Chowdhury
The servers that we get these occasional
errors on, are mostly dual AMD boxes (MP 1800s, typically) running 2.4.18. Are the decoding rules, present in Intel's IA-32 MCA documentation, applicable
No idea. I've not seen an AMD document specifying how they encode their error reports. Normally its overheating and/or crappy power supplies. The dual Athlons burn power and output huge amounts of heat.