[net.periphs] Pitfalls of current SASI/SCSI controllers

gnu@sun.uucp (John Gilmore) (02/20/85)

> | Is it possible to find SCSI disk and tape controllers with reasonable
> | performance?  (Defined as delivering a useful fraction of the potential
> | speed of the device, i.e. not requiring 5:1 interleaving on the disk
> | because the stupid controller can't keep up.)...
> 
> Surpisingly, 1:1 is possible. While earlier SASI/SCSI controllers were all of
> the single-buffer type, newer (and still low-cost) boards from Adaptec, OMTI,
> and (maybe) Western Digital are all capable of 1:1 operation (they look like
> FIFOs).

Note that most, if not all, SASI/SCSI disk controllers are too smart
for YOUR own good.  They "handle" bad block mapping but do a lousy job
of it and won't let your software improve their job.  For example, the
Adaptec we use is utterly reliable and runs 1:1 interleaf, but:

     *	Won't let you reformat part of the disk -- you have to do it all.
     *	Won't let you remap a sector or track if it goes bad during
	normal operation -- you have to reformat the whole disk.
	(After dumping any data you want to keep, assuming your dumping
	tools make it past the bad block...)
     *	Won't do automatic error checking & recovery.
	(Controller won't do the job itself either.)
     *	Won't tell you where cylinder boundaries are so you can use
	fancy 4.2BSD code to optimize disk layout (only partly true --
	it will tell you but you have to scan the disk to find out)

All our SMD controllers give us full access to these facilities -- and
we use them.

We are looking for a controller (or Adaptec ucode upgrade) that will let us
reformat & map sectors or tracks, or do slip sectoring.