Subject: Re: I/O priorities
To: None <tech-kern@netbsd.org>
From: der Mouse <mouse@Rodents.Montreal.QC.CA>
List: tech-kern
Date: 06/21/2002 13:47:20
> At a coarse level, the symptoms sound afwully like scenarios in
> standard OS textbooks: simplistic seek optimization causes
> starvation.  The textbook approach is to do SCAN elevator sorts: once
> the disk arm starts a sweep, new requests are postponed until the
> next, reverse sweep.  Within a sweep, requests are serviced in-order.

And of course this is complicated by ordering constraints between
various writes, which I seldom see mentioned in this connection.  Such
constraints mean either deferring some requests until `later', cutting
a scan short, or doing something a good deal more complicated than the
simple scan algorithm outlined....

/~\ The ASCII				der Mouse
\ / Ribbon Campaign
 X  Against HTML	       mouse@rodents.montreal.qc.ca
/ \ Email!	     7D C8 61 52 5D E7 2D 39  4E F1 31 3E E8 B3 27 4B