From owner-freebsd-hackers  Thu Sep 28 10:43:58 1995
Return-Path: owner-hackers
Received: (from root@localhost)
          by freefall.freebsd.org (8.6.12/8.6.6) id KAA22038
          for hackers-outgoing; Thu, 28 Sep 1995 10:43:58 -0700
Received: from phaeton.artisoft.com (phaeton.Artisoft.COM [198.17.250.211])
          by freefall.freebsd.org (8.6.12/8.6.6) with ESMTP id KAA22021
          for <hackers@FreeBSD.ORG>; Thu, 28 Sep 1995 10:43:55 -0700
Received: (from terry@localhost) by phaeton.artisoft.com (8.6.11/8.6.9) id KAA12846; Thu, 28 Sep 1995 10:38:08 -0700
From: Terry Lambert <terry@lambert.org>
Message-Id: <199509281738.KAA12846@phaeton.artisoft.com>
Subject: Re: News Server VM & I/O issues (was: Re: Big win for BSD/OS compatibility)
To: jgreco@brasil.moneng.mei.com (Joe Greco)
Date: Thu, 28 Sep 1995 10:38:08 -0700 (MST)
Cc: taob@io.org, hackers@FreeBSD.ORG
In-Reply-To: <199509280600.BAA05493@brasil.moneng.mei.com> from "Joe Greco" at Sep 28, 95 01:00:39 am
X-Mailer: ELM [version 2.4 PL24]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Content-Length: 1059      
Sender: owner-hackers@FreeBSD.ORG
Precedence: bulk

> Rod made an interesting suggestion:
> 
> > Start chasing around in the UFS code and see if you can some how create
> > seperate priorities of importanace for meta data vs file block data in
> > the VM cache, in your application tossing out meta data is VERY expensive.
> > A few quick emails to John Dyson or David Greenman might point you to
> > some tuneable parameters in the system that would help you out.

This is actually bogus; it violates the locality rules that cause a
cache to function in the first place.

One soloution would be to make writes be ordered but not synchronus.  The
synchronus metadata updates are a poor mechanism for guaranteeing the
order is (1) metadata and (2) data.

Unfortunately, this approach is currently patent-pending by USL, although
I believe there is sufficient prior art that it isn't really patentable,
that usually has little bearing on whether or not a patent is granted.


					Terry Lambert
					terry@lambert.org
---
Any opinions in this posting are my own and not those of my present
or previous employers.