[sclug] Multiple drives

[Neil Haughton]

> This could well be wrong, but...
>
> HDDs usually support a faster IDE mode than optical drives, and if I
> recall correctly, the fastest IDE mode on a single channel is
> constrained by the slowest device on the channel. So by putting an
> optical drive on each channel, you're restricting the speed of the hard
> drives.
>   

Apparently that's not quite all there is to it. My research today 
suggests that 'modern' IDE chipsets will allow both ATA devices on a 
channel to operate at full speed, even if they are different speeds, but 
a mix of ATA and ATAPI drives on a channel will cause the data transfer 
to be at the ATAPI (slower) speed if both devices are in use. Apparently 
if the ATAPI drive isn't being used (ie read from or written to) the ATA 
device will still operate at its full speed.

Hang the cost  ;-) I'm going to put in another 2xIDE card, and run the 
HDDs each as masters on IDE0 and IDE1, and the optical drives as masters 
on the two new channels.  I trust Linux will be happy with this arrangement?

Neil.

> What I'd suggest is trying some benchmarks. 'hdparm -t -T /dev/hd?' will
>  measure the transfer speed of a drive. (It'll give you two numbers, one
> the speed the drive's hardware can push data over the cable, and the
> other the speed it can read data off the disk.) If you try this in
> various combinations, you should be able to figure out the best combination.
>
> - --
> +- David Given --McQ-+
> |  dg at cowlark.com    | "Those that repeat truisms, are also forced to
> | (dg at tao-group.com) | repeat them." --- Anonymous from Slashdot
> +- www.cowlark.com --+
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.1 (GNU/Linux)
> Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org
>
> iD8DBQFEK7YCf9E0noFvlzgRAuioAJ4/PJp4UG6FGQ2ubmh8QccgET9V9wCaAxVV
> ZyMRjOocyz9uZHaFVDaDO0Y=
> =xC1+
> -----END PGP SIGNATURE-----
>   
> ------------------------------------------------------------------------
>
> _______________________________________________
> sclug mailing list
> sclug at sclug.org.uk
> http://www.sclug.org.uk/mailman/listinfo/sclug
>