On Wed, Feb 19, 2003 at 11:00:58AM -0600, Greg Copeland wrote:
> On Wed, 2003-02-19 at 10:31, Jonathan Bartlett wrote:
> > Actually, IDE has performance issues even with only 1 disk per channel.
> > The SCSI command set allows disconnected operations, so that the kernel
> > can send commands (get block xxx, get block yyy, get block zzz) and keep
> > sending commands while the drive processes the answers. With IDE, it is a
> > synchronous transmission - get blox xxx, wait until disk processes,
> > receive block xxx, get block yyy, wait until disk processes, receive block
> > yyy. SCSI disks can also reorder the requests and service them based on
> > how quickly it can get to each one.
>
> Agreed. I think we are saying the same thing. You just happen to go
> into more detail. :P My point being, if you use IDE, you should be
> attempting to use a disk per channel. BTW, on a different note, IBM
> created some IDE drives which allow for command tagging which allows for
> multiple outstanding IDE commands, however, I have no idea what the
> availability of said drives and drivers are. I'm actually fairly
> ignorant on the exact device details. You wouldn't happen to have the
> skinny of those things would ya? They still being made?
For reference it's called Tagged Command Queueing (TCQ). It appears to be
supported by the IBM Deskstar drive only so far. There is some support in
Linux 2.5 AFAICS. It doesn't work with all controllers either apparently.
Bit of a mish-mash really.
http://lwn.net/2002/0411/a/ah-tcq.php3
--
Martijn van Oosterhout <kleptog@svana.org> http://svana.org/kleptog/
> Support bacteria! They're the only culture some people have.