6 Nov
1999
6 Nov
'99
8:42 a.m.
Anyone have any experience with the Dilog DQ696 ESDI
disk controllers? I
just installed one in a MicroVAX III and I've been getting really _dismal_
performance out of it. I had it misterminated for a bit (the middle drive
(of two) was terminated rather than the end drive). But I fixed that
without any increase in performance. The format/block analysis took about
10 hours for a 600MB disk.
The interleave was set to 1 but that seems pretty standard these days,
should I up it to 2 or 3? or perhaps reduce it to 0?
The drives are Micropolis 1568's so they're nice fast drives.
Were these 1568's used previously on this Dilog controller? If not, they're
probably configured for use on a PC-clone, and have the sector length set
on the "short" side. The PLO in the Dilog isn't locking onto successive
blocks, and as a result it's reading only one sector per revolution.
The cure is to set the sector length on the 1568's to be a bit longer.
You'll lose a little bit of capacity (and will have to reformat), but it'll
be able to read more than one sector per revolution.
Setting the interleave can interact with this, of course. If you don't
know how to alter the block length on your drives, try altering the
interleave and see if that helps any.
How about
Q-bus priority?
No, if things are slow with the controller running the built-in exerciser
(which *don't* move disk data over the Q-bus) it's not a Q-bus problem.
It's a drive problem.
--
Tim Shoppa Email: shoppa(a)trailing-edge.com
Trailing Edge Technology WWW: http://www.trailing-edge.com/
7328 Bradley Blvd Voice: 301-767-5917
Bethesda, MD, USA 20817 Fax: 301-767-5927