RE: CPU Utilization

From: S.J.Haribabu (sjhari_at_microsoft.com)
Date: 06/04/04


Date: Fri, 04 Jun 2004 20:26:18 GMT

Hi,

A processor bottleneck occurs when a very high percentage of the CPU is
being utilized fairly often. You can expect CPU use to spike—that's a
normal, quick response to an event. But if a CPU consistently runs at
approximately 90 percent or above, that indicates a processor problem. In
this case, there are a number of actions you can take:

• Upgrading the CPU is usually a good idea.
 
• Additional CPUs can help.
 
• Move some of the processors off that machine onto another one, or
implement network load balancing policies. (See the "Network Load
Balancing" module for more information.) By spreading content across a
number of servers, you are in effect spreading out the impact of the hits
to all those servers.
 

Note: As you add more processors and more RAM, don't forget about your
level 2 cache. The L2 cache should grow as the amount of RAM increases, or
you may not realize the performance gains you expect.

For more information look at
http://www.microsoft.com/technet/prodtechnol/windows2000serv/evaluate/w2khos
t/w2ktpt.mspx

Also look for memory management, Disk optimization, etc.

Thanks,

sjhari@online.microsoft.com

This posting is provided "AS IS" with no warranties, and confers no rights.

 



Relevant Pages

  • Re: Optimization problem
    ... CPU core and is able to respond to the CPU at full speed (i.e. read or ... each of the four CPU cores has 32 KBytes of L1 cache for data (and ... When you go out of L2 cache as well, the CPU must rely on the plain RAM. ... harddisk) is really called _paging_. ...
    (comp.lang.java.programmer)
  • IPv6 oops on ifup in latest BK
    ... Using ACPI for SMP configuration information ... Initializing CPU#0 ... CPU: Trace cache: 12K uops, ...
    (Linux-Kernel)
  • Re: Purchasing the correct hardware: dual-core intel? Big cache?
    ... to 1/2 day purely by tuning the SQL. ... disk/data layout, then finally the CPU. ... Fact is, with 4G of RAM most of the data sits in RAM, so reads incur ... but I'm not sure how to tell if the cache is getting used ...
    (freebsd-questions)
  • SCSI CDROM issue in kernels >= 2.6.14-rc3
    ... CPU: Trace cache: 12K uops, ... MEM window: disabled. ... SCSI device sda: 17928698 512-byte hdwr sectors ...
    (Linux-Kernel)
  • mptscsih: ioc1: attempting task abort! (sc=d6e8a980)
    ... CPU 2: Machine Check Exception: 0000000000000004 ... OEM ID: INTEL Product ID: Bridge CRB APIC at: 0xFEE00000 ... CPU: Trace cache: 12K uops, ... SCSI device sda: 287132440 512-byte hdwr sectors ...
    (Linux-Kernel)