Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site astrovax.UUCP
Path: utzoo!linus!decvax!harpo!eagle!mhuxl!ulysses!princeton!astrovax!wls
From: wls@astrovax.UUCP
Newsgroups: net.unix-wizards
Subject: Re: m and n on 750 w/ RA81's
Message-ID: <154@astrovax.UUCP>
Date: Wed, 30-Nov-83 15:20:14 EST
Article-I.D.: astrovax.154
Posted: Wed Nov 30 15:20:14 1983
Date-Received: Fri, 2-Dec-83 00:29:29 EST
References: <169@sfucmpt.UUCP>
Organization: Princeton Univ. Astrophysics
Lines: 15

One would expect that the elapsed time would be sensitive to "m" but insensitive
to "n".  "m" is the interlace factor, the spacing of the blocks on the disk,
which sets how much time there is before the the system must be ready to access
the next block.  Otherwise the system has to wait for the disk to go all the way
around and catch it on the next revolution. 
  The "n" value, in contrast, is provided only to keep the interlacing within
a cylinder, so that no head movement has to take place until that cylinder is
finished.  If "n" is wrong then the interlacing straddles cylinder boundaries
and extra head movement does take place.  However, for reasonable values of
"m" and "n" the extra number of disk accesses which involve head movement is
still small compared to the number of accesses which do not, and the effect on
the elapsed time is only minor.

By the way, I hope someone at Berkeley is saving this m and n info to include
in future manuals.
-- 
Bill Sebok			Princeton University, Astrophysics
{allegra,akgua,burl,cbosgd,decvax,ihnp4,kpno,princeton}!astrovax!wls