Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 30 Apr 2001 17:35:56 -0700
From:      "'Alfred Perlstein'" <bright@wintelcom.net>
To:        Charles Randall <crandall@matchlogic.com>
Cc:        =?iso-8859-1?Q?=AAL=AD^=B6W?= <r88074@csie.ntu.edu.tw>, Freebsd-Hackers <freebsd-hackers@FreeBSD.ORG>
Subject:   Re: write() vs aio_write()
Message-ID:  <20010430173555.H18676@fw.wintelcom.net>
In-Reply-To: <5FE9B713CCCDD311A03400508B8B30130828EC88@bdr-xcln.corp.matchlogic.com>; from crandall@matchlogic.com on Mon, Apr 30, 2001 at 11:24:42AM -0600
References:  <5FE9B713CCCDD311A03400508B8B30130828EC88@bdr-xcln.corp.matchlogic.com>

next in thread | previous in thread | raw e-mail | index | archive | help
* Charles Randall <crandall@matchlogic.com> [010430 10:26] wrote:
> Regarding aio_*, Alfred Perlstein writes:
> >It's a good idea to use it for disk IO, probably not a good
> >idea for network IO.
> 
> Could you elaborate?

Sure.

Network IO can be done without blocking (unless you take a fault
on the source address of your data).  Hence the additional context
switching required by aio is not needed.

Disk IO probably stands a good chance of blocking your application,
if you can offload that blocking to a kernel thread you should be
able to continue serving content.

-- 
-Alfred Perlstein - [alfred@freebsd.org]
Instead of asking why a piece of software is using "1970s technology,"
start asking why software is ignoring 30 years of accumulated wisdom.

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-hackers" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20010430173555.H18676>