tlug: Bus master IDE performance under Linux

["Francis Brian O'Carroll" <ocarroll@example.com>]

--------------------------------------------------------
tlug note from "Francis Brian O'Carroll" <ocarroll@example.com>
--------------------------------------------------------
How well do current versions of linux hande "bus master ide" disks?
Is it a cheaper alternative to scsi with the same performance, or not?

One of the researchers here did some tests with netbsd (he probably
had to do some kernel hacking since he mentioned having to consult the
disk controller chip's spec sheet).

He used 'time' to measure the execution of a busy loop process with no
other user processes running, and then timed the busy loop when a dd
was in progress.

I won't quote exact numbers, but under programmed i/o the busy loop
took > 5 times as long to run when there was a dd running in parallel
compared to when the busy loop was the only process.  Under bus master,
the busy loop only took 5% longer to run with a dd in parallel.

Sustained disk performance was 5MB/s in both cases. I know that the
peak performance of bus master ide is often quoted as 16 MB/s :-)

And speaking of benchmarking, are there any system benchmarks for
Linux machines apart from BogoMips?

Frank O'Carroll
Tsukuba & Tokyo
-----------------------------------------------------------------
a word from the sponsor will appear below
-----------------------------------------------------------------
The TLUG mailing list is proudly sponsored by TWICS - Japan's First
Public-Access Internet System.  Now offering 20,000 yen/year flat
rate Internet access with no time charges.  Full line of corporate
Internet and intranet products are available.   info@example.com
Tel: 03-3351-5977   Fax: 03-3353-6096