Subject: Re: Disk scheduling policy (Re: NEW_BUFQ_STRATEGY)
To: Matt Thomas <firstname.lastname@example.org>
From: Thor Lancelot Simon <email@example.com>
Date: 12/01/2003 18:57:18
On Mon, Dec 01, 2003 at 03:47:44PM -0800, Matt Thomas wrote:
> At 03:16 PM 12/1/2003, Jason Thorpe wrote:
> >-----BEGIN PGP SIGNED MESSAGE-----
> >Hash: SHA1
> >On Dec 1, 2003, at 3:11 PM, Thor Lancelot Simon wrote:
> My assumption was that you *always* drain the current queue before
> switching to other queue. In that scenario, there is no starvation.
The way I read the SGI text, they do N requests from queue A, then
N requests from queue B, and so forth. A simple implementation of
this seems like it might disrupt the elevator sort quite badly, so I
wonder if they actually did something more clever.
I can think of a reasonably simple way to tune "N" if we can
characterize the seek time of the disk... but that seems dangerous.
Thor Lancelot Simon firstname.lastname@example.org
But as he knew no bad language, he had called him all the names of common
objects that he could think of, and had screamed: "You lamp! You towel! You
plate!" and so on. --Sigmund Freud