|
[Date Prev] [Date Next] [Thread Prev] [Thread Next] [Date Index] [Thread Index]
Bill Jonas wrote:
>
> On Tue, May 28, 2002 at 10:11:58AM -0600, W. Chris Shank wrote:
> > so there is no way of doing this to increase your write speed?
>
> Buy a disk that spins faster. :)
Doesn't that depend on the hardware? For IDE-Raid in software, where
you can only talk to one of the IDE disks on an interface at a time, I'd
agree. (You get around that by only using one disk per IDE
interface...) But for SCSI-Raid, certainly with a hardware controller,
and possibly with software raid, the wait time for each 'stripe' doesn't
change. But because you have them distributed across 2 drives, the
overall time to write them can be up to twice as fast. Of course,
that's the theoretical limit. You won't see that in real life.
I've got some test results from a SCSI-hardware RAID setup using
striping:
For a single disk:
---Sequential Output (nosync)--- ---Sequential Input--
--Rnd Seek-
-Per Char- --Block--- -Rewrite-- -Per Char- --Block---
--04k (03)-
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU
/sec %CPU
octy 1* 500 13774 98.1 31450 26.2 5259 3.1 8435 53.5 46774 19.0
709.3 5.7
For a 4 disk stripe (RAID 0):
---Sequential Output (nosync)--- ---Sequential Input--
--Rnd Seek-
-Per Char- --Block--- -Rewrite-- -Per Char- --Block---
--04k (03)-
Machine MB K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU K/sec %CPU
/sec %CPU
linux 1* 500 15180 99.9 83130 34.3 8514 4.3 13314 82.3 77350 22.7
1451.6 5.4
--
Kevin Brosius
______________________________________________________________________
Philadelphia Linux Users Group - http://www.phillylinux.org
Announcements-http://lists.phillylinux.org/mail/listinfo/plug-announce
General Discussion - http://lists.phillylinux.org/mail/listinfo/plug
|
|