[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]

Re: [K12OSN] Linux "Software RAID"

On Fri, 2008-08-08 at 07:37 -0400, Rob Owens wrote:

> Realistically, a hardware RAID costs about $300, and that could buy me a 
> 2nd 4-core processor.  I have to believe that a modern 4-core processor 
> is way more capable of handling a RAID rebuild than whatever chip is 
> onboard a hardware RAID controller.  Of course, I haven't done any 
> testing...

Most of the hardware RAID devices I have read full specs on use a
PowerPC chip for their brain so they have some horsepower behind their

The overhead for software RAID will depend very heavily on how the
system is built and configured. A crappy chipset will give crappy
performance. A chipset that can parallel-pipeline data through to
multiple PCI busses will smoke the charts. But those chipsets will only
be on the high-end server-class boards. But those are the boards you
want to be using anyway. :-)

Bear in mind that the real performance hit happens on the write. During
a read you can actually get a speedup using multiple spindles (on
multiple channels - single channel with multi-drive will be the speed of
one drive). This holds true for both hardware and software RAID. The
hardware RAID will be then limited by the bus speed of the slot it's
plugged into. Software RAID might outperform here if the chipset is good
and provides multiple lanes for sata socket traffic. Few boards do this.

For LTSP use, with heavily shared hard drive space on a 100-seat server,
the cpu's will be taxed at 100% load all the time. Go hardware on this
set-up. If you have a classroom server with a single-chip, dual core
athalon, go software.
James P. Kinney III          
CEO & Director of Engineering 
Local Net Solutions,LLC                           

GPG ID: 829C6CA7 James P. Kinney III (M.S. Physics)
<jkinney localnetsolutions com>
Fingerprint = 3C9E 6366 54FC A3FE BA4D 0659 6190 ADC3 829C 6CA7

This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

[Date Prev][Date Next]   [Thread Prev][Thread Next]   [Thread Index] [Date Index] [Author Index]