RAID-3?

[Poul-Henning Kamp]

In message <41244F17.9030007 at samsco.org>, Scott Long writes:
>Greg 'groggy' Lehey wrote:

>> I can see that as a great advantage, but it's not part of the RAID-3
>> definition, and I can't see why you couldn't expand RAID-5 in a
>> similar manner.  Am I missing something?
>> 
>> Greg

Yes you are missing the complexity of the code to implement it.
As far as I know, RaidFrame is the only working implementation of
RAID5 with two redundant disks.

>Yes, you are!  The advantage of RAID-3 is that there are NO
>Read-Modify-Write cycles when writing blocks.  Period.  Zippo.  None.
>Every write takes exactly the same amount of time.  There is no waiting
>for data to be read off of any disks.  That is why it's nice to
>applications that require fixed latency.  RAID-3 has no concept of
>stripe sizes becuase of this, unlike 4 and 5.
>
>Scott

Well, in RAID3 the stripe size becomes your sectorsize and if you are
using a filesystem that demands a particular sectorsize you may be
prevented from using RAID3 because of that.  UFS/FFS does not have
this problem.

The other thing is that RAID5 can be made in any configutation from
1+1 to N+1, whereas RAID3 is generally limited to 2^N+X

Poul-Henning

-- 
Poul-Henning Kamp       | UNIX since Zilog Zeus 3.20
phk at FreeBSD.ORG         | TCP/IP since RFC 956
FreeBSD committer       | BSD since 4.3-tahoe    
Never attribute to malice what can adequately be explained by incompetence.