🚀 go-pugleaf

On Fri, 1 Oct 1999 17:06:41 -0700, "Ron Reaugh"
<Ron-Reaugh@worldnet.att.net> caused to appear as if it was written:

>>I was aware of the need to calculate parity during the write in RAID 5 but
>>I thought the ability to write a small peice of the whole file to each disk
>>(in RAID 5) would be more significant timewise.

>You are correct if the RISC onboard the HW RAID 5 controller is fast enough.

Generally speaking, the RAID controller is irrelevant to anyone but the
system designer (be it RISC or anything else: plenty of monster RAID systems
use CISC CPUs).

>>Does RAID 5 ever get better than RAID 1,  as you add disks?

>At 3 disks RAID 5 be faster IF the RISC is fast.

Untrue.

>>ie RAID 1 vs   a  8  Disk RAID 5
>>is RAID 1 still better?
>
>
>RAID 1 likely wont be faster even at 2 disks if the RISC is fast.

Nonsense.

And confused: a 2 disk RAID5 is identical to very ineffient mirroring...

Assuming you mean a 2+1 RAID5 array, you are only correct IF the RAID1
controller is poorly designed (which, sad to say, means that you are correct
a lot of the time!).

But a *smart* controller splits incoming requests into lists of requests for
each disk pair, and issues one command to each side of the mirror.  So a 2+2
RAID1 set (striped or concatenated) can handle 4 simultaneous reads, while
the equivalent 2+1 RAID5 set can only handle 3...

Malc.