[mythtv-users] The Bigger... Disk contest, Fall 2007 edition

Jay R. Ashworth jra at baylink.com
Sat Oct 20 00:28:11 UTC 2007


On Fri, Oct 19, 2007 at 10:45:36AM -0600, Brian Wood wrote:
> I discovered that virtually all of my SCSI termination problems went
> away when I went to active termination, as opposed to passive resistors.
> 
> Using good quality cables also makes a difference, as does staying away
> from devices that use DB-25s for SCSI. You want a cable that has a
> separate ground conductor for each data line, not one common one for all
> of them.
> 
> Using the same type of cable for each link in an external chain is also
> a good idea, nor some hodgepodge from your junk drawer.
> 
> Going to low-voltage differential units, instead of the earlier
> single-ended ones, also helps, especially at high speeds or with long
> cables.
> 
> Like anything else SCSI works well if you do it correctly and don't try
> and cheap out, but ATA type interfaces are certainly in more common use
> for home systems. PCs are marketed based on CPU speed because most
> buyers of such machines do not know or even care that the rest of the
> machine can't come close to keeping up with the CPU, while commercial
> servers are usually rated based on benchmark suites that test all
> aspects of the hardware.

The Three Laws Of SCSI:

1) It's the termination, stupid.

2) It's the cable, stupid.

3) It's the jumpers, stupid.

Apply in whichever order seems best; old work, top to bottom, new work,
bottom to top.

Cheers,
-- jra
-- 
Jay R. Ashworth                   Baylink                      jra at baylink.com
Designer                     The Things I Think                       RFC 2100
Ashworth & Associates     http://baylink.pitas.com                     '87 e24
St Petersburg FL USA      http://photo.imageinc.us             +1 727 647 1274


More information about the mythtv-users mailing list