RedHat 8.0 finds too many cpu's

[MWillo]

I apologise in advance for the rather lengthy question

After installing RedHat Professional 8.0 on a Compaq Proliant ML370 with dual Xeon processors, we have a problem where the server will stay up for about 2-3 hours and then freeze completely.

Version info
Red Hat Linux release 8.0 (Psyche)
Kernel 2.4.18-14smp on an i686

In the logs it seems that Linux thinks the machine has 4 cpu's and this is backed up by /proc/cpuinfo which is as follows:
processor : 0
vendor_id : GenuineIntel
cpu family : 15
model : 2
model name : Intel(R) Xeon(TM) CPU 2.80GHz
stepping : 7
cpu MHz : 2785.067
cache size : 512 KB
Physical processor ID : 0
Number of siblings : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
bogomips : 5522.26

processor : 1
vendor_id : GenuineIntel
cpu family : 15
model : 2
model name : Intel(R) Xeon(TM) CPU 2.80GHz
stepping : 7
cpu MHz : 2785.067
cache size : 512 KB
Physical processor ID : 0
Number of siblings : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
bogomips : 5556.67

processor : 2
vendor_id : GenuineIntel
cpu family : 15
model : 2
model name : Intel(R) Xeon(TM) CPU 2.80GHz
stepping : 7
cpu MHz : 2785.067
cache size : 512 KB
Physical processor ID : 3
Number of siblings : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
bogomips : 5556.67

processor : 3
vendor_id : GenuineIntel
cpu family : 15
model : 2
model name : Intel(R) Xeon(TM) CPU 2.80GHz
stepping : 7
cpu MHz : 2785.067
cache size : 512 KB
Physical processor ID : 3
Number of siblings : 2
fdiv_bug : no
hlt_bug : no
f00f_bug : no
coma_bug : no
fpu : yes
fpu_exception : yes
cpuid level : 2
wp : yes
flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat
pse36 clflush dts acpi mmx fxsr sse sse2 ss ht tm
bogomips : 5556.67

The logs also report the following error which usually indicates an issue with inconsistant stepping on the processors which is as you can see from above not the case.

TRR - CPU's inconsistent fixed MTRR settings.

I realise this could well be a hardware/bios issue with the server, but any light anyone could shed on it would be much appreciated.

 

[Riaan]

hi there MWillo

- 2 or 4 CPU's?
According to ZDNET: Hyperthreading--a performance-enhancing technology that lets one chip act something like two

http://zdnet.com.com/2100-1103-963942.html

your brand new server has HyperThreading technology in the CPU's, in which the amount of physycal CPU's are doubled to get the amount of logical CPU's. You should be able to disable it in the BIOS, which would let Linux see just 2 CPU's again.

the Intel's site has more

http://www.intel.com/technology/hyperthread/

- inconsistent fixed MTRR settings.
You can safely ignore this message.:
From the search results on Google, the following seems the best explanation

http://www.faqchest.com/linux/KERNEL/kern-99/kern-9911/kern-991180/kern99113003_06308.html

+++++
This means that the BIOS didn't set up the MTRRs (Memory Type Range
Registers) in the same way for both CPUs.

This is actually allowed in the MP BIOS definition. But it's kind of a
rotten trick to play on someone if their OS doesn't know how to take
care of it, as it usually ends up just making your system run pretty
slow, but otherwise functionally correct.

So, the OS is just warning you about it. I forget offhand, but suspect
that Linux fixes this by setting the other CPUs' MTRRs to the same as
the boot processor, which is the Right Thing to do.
++++++

- Server locking up
I would suggest the following:
a) Use the smart start Boot CD with diagnostic utilities to diagnose and verify your server
b) Install the Compaq health and welness driver and agents. You can access the Integrated Management Log and agent data via a web browser, which will show if a component is degraded/failed

Here is the addres for Proliant and Linux

http://h18000.www1.hp.com/products/servers/linux/

It has links to Software (agents, etc) and Drivers as well as Documentation to get the software going
Riaan van Niekerk
Unix sysadmin & groupie
Potchefstroom University
South Africa

 

[MWillo]

Thanks for your input,

I'll let you know the outcome