[techtalk] Re: Benchmark programs
- From: Jack <jyork@xxxxxxxxxxx>
- To: techtalk@xxxxxxxxxxxxx
- Date: Sat, 6 Mar 2004 09:02:31 -0500
Hello Nick,
> Just to give you some insitie. Raid controlers differ in speeds based on
> the interface ie. SCSI, (parallel, serial) ide. It is unlikely that an
> onboard (depending on manufacturer) will difer from an "ad-in" card in terms
> of speed. You will, most certanly notice a performance increase with a
> stripe. The sole premase of the idea states that if one drive is active the
> controler will access the other drive. Obviously, the more drives you have
> on the array the faster the system. To answer your question I would try 3d
> Mark 2000 for starters. For a more difinative answer YOU are the best judge
> of how you system responds. Try it both ways. I run striped arrays on all my
> machines, because I know it's faster. If redundance is what you seek,
> consider another drive (out of the array). If your running NTFS it can be a
> trifling experience to rebuld an array with a mirror if one drive fails. So,
> my point is why waste the space.
I got a deal on the drives, two 200 gig WD's, so wasting space is not
really a concern. I'm not as diligent as I should be with backups
though. That's why I thought the mirror would be best for me. I'll
give the test routine a try. I've also ran across a few that test
disk speed only though they are older so I'm concerned that they may
cause trouble with NTFS if I try using the write portions so I haven't
done that yet. Guess I'll just have to try the raid both ways and see
what I like best. Thanks for the input.
--
Jack
http://www.mycandysupplier.com
Other related posts:
