RAID-3?

John-Mark Gurney gurney_j at resnet.uoregon.edu
Wed Aug 18 23:44:06 PDT 2004


Greg 'groggy' Lehey wrote this message on Thu, Aug 19, 2004 at 16:08 +0930:
> On Wednesday, 18 August 2004 at 23:28:48 -0700, John-Mark Gurney wrote:
> > Greg 'groggy' Lehey wrote this message on Thu, Aug 19, 2004 at 15:52 +0930:
> >>> Your quoted text also seems a bit subjective as there are very valid
> >>> reasons for RAID-3, especially if one is looking for consistent
> >>> low-latency transactions like in video recorders and servers.
> >>
> >> Well, I did use *exactly* this example.  I also pointed out that the
> >> relative performance of modern disk subsystems is adequate for a
> >> single streaming video channel.
> >>
> >> Low latency depends on the number of concurrent accesses.  RAID-3
> >> handles concurrent access poorly, exactly because it accesses all
> >> disks for each transfer.
> >
> > One thing that RAID-3 has is that you never have to do a READ/MODIFY
> > cycle when you do writes.  Until we implement a write-through cache
> > geom module, raid-5 will continue to substandard performance.
> 
> Even then, RAID-5 might have higher bandwidth under some
> circumstances.

Pick your tool, and you can always find a good example and a bad
example of how to use the tool.  Doesn't mean it's bad.

> My real question about RAID-3 remains: what use is it?  This isn't
> nit-picking, it's certainly not a criticism of pjd.  I just don't see
> any practical use on FreeBSD machines.

I originaly was working on a RAID-3 module (which is possibly where pjd
got his idea) that used Luigi's FEC code.  The advantage of this code was
the fact that you could have n parity disks beyond the m data disks.  The
advantage of this was that you could loose any n disks, and your data is
still recoverable.  Unlike with RAID-4/5 implementations where if you
happen to loose a second disk (due to a power surge or something) while
rebuilding, you'd be SOL.  That type of redundancy is good thing to have.

-- 
  John-Mark Gurney				Voice: +1 415 225 5579

     "All that I will do, has been done, All that I have, has not."


More information about the cvs-all mailing list