Multiple choices for RAID-0, best performance?

[Paul Mather]

On Fri, 25 Jun 2004 10:23:37 +0200, S?ren Schmidt<sos at DeepCore.dk> wrote:

> Brad Knowles wrote:
> 
> >     My understanding is that regular vinum won't work on 5.x, you need 
> > geom-vinum for that.  Recent postings from Sren have lead me to believe 
> > that none of the ataraid stuff works with 5.x, either -- you can use 
> > them as plain ATA controllers, but the RAID stuff has not yet been 
> > worked out.
> 
> Wrong, ataraid works fine on 5.x.

Ataraid has never worked entirely for me under 5.x.  Specifically, I
have *never* been able to get "atacontrol rebuild" to kick off
reconstruction after replacing a drive in a DEGRADED array.  This has
been on an assortment of standard ATA controllers.  The most recent on
which I attempted it is this one:

atapci0: <Intel PIIX4 UDMA33 controller> port 0xf000-0xf00f,0x376,0x170-0x177,0x3f6,0x1f0-0x1f7 at device 7.1 on pci0

atapci0 at pci0:7:1:       class=0x010180 card=0x00000000 chip=0x71118086 rev=0x01 hdr=0x00
    vendor   = 'Intel Corporation'
    device   = '82371AB/EB/MB PIIX4/4E/4M IDE Controller'
    class    = mass storage
    subclass = ATA

Naturally, not being able to rebuild an array in the event of drive
failures makes ataraid useless for my purposes, so I gave up on it.

> However there are some ATA software RAID setups that I havn't done the 
> metadata code for yet, but they'll get there eventually time permitting. 
> You can always use atacontrol to setup a RAID on *any* ATA 
> disk/controller setup, however you will need a BIOS with a known 
> metadata layout (Promise/Highpoint) to be able to boot from it (se man 
> atacontrol(1) for details)..

You should, then, clarify the man page for atacontrol, as it appears to
conflict with the above:

     create   Create a type ATA RAID.  The type can be RAID0 (stripe), RAID1
              (mirror), RAID0+1 or SPAN (JBOD).  In case the RAID has a RAID0
              component, the interleave must be specified in number of sec-
              tors.  The RAID will be created of the individual disks named
              disk0 ... diskN.

              Although the ATA driver allows for creating an ATA RAID on disks
              with any controller, there are restrictions.  It is only possi-
              ble to boot on an array if it is either located on a ``real''
              ATA RAID controller like the Promise or Highpoint controllers,
              or if the RAID declared is of RAID1 or SPAN type; in case of a
              SPAN, the partition to boot must reside on the first disk in the

In other words, the man page says that it is possible to boot without
the need for a RAID BIOS (known or otherwise) so long as the ATA RAID is
of type RAID1 or SPAN (with further restrictions).

Whilst on the subject of clarifying man pages, is the following entry in
"man atacontrol" literally correct:

     rebuild  Rebuild a RAID1 array on a RAID capable ATA controller.

Specifically, does it mean that if you don't have a RAID-capable ATA
controller (i.e., you have only a standard ATA controller) then
"atacontrol rebuild" will not work?  (If that is the case, it is strange
to allow certain ataraid arrays to be created and booted on non-RAID
controllers but not rebuilt.)

Cheers,

Paul.
-- 
e-mail: paul at gromit.dlib.vt.edu

"Without music to decorate it, time is just a bunch of boring production
 deadlines or dates by which bills must be paid."
        --- Frank Vincent Zappa