*BSD News Article 34510


Return to BSD News archive

Xref: sserve comp.os.386bsd.questions:12455 comp.os.386bsd.misc:3260
Newsgroups: comp.os.386bsd.questions,comp.os.386bsd.misc
Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!bunyip.cc.uq.oz.au!munnari.oz.au!constellation!news.uoknor.edu!ns1.nodak.edu!netnews.nwnet.net!pnl-oracle!osi-east2.es.net!lll-winken.llnl.gov!sol.ctr.columbia.edu!howland.reston.ans.net!math.ohio-state.edu!jussieu.fr!univ-lyon1.fr!swidir.switch.ch!newsfeed.ACO.net!Austria.EU.net!EU.net!uunet!brunix!mhw
From: mhw@cs.brown.edu (Mark Weaver)
Subject: Re: Whats wrong with Linux networking ???
Message-ID: <1994Aug15.175932.10899@cs.brown.edu>
Sender: news@cs.brown.edu
Organization: Brown University Department of Computer Science
References: <RSANDERS.94Aug9003813@hrothgar.mindspring.com> <32ll52$n7d@quagga.ru.ac.za> <1994Aug15.034939.20997@cs.brown.edu> <michaelv.776931077@ponderous.cc.iastate.edu>
Date: Mon, 15 Aug 1994 17:59:32 GMT
Lines: 34

In article <michaelv.776931077@ponderous.cc.iastate.edu>,
Michael L. VanLoon <michaelv@iastate.edu> wrote:
>In <1994Aug15.034939.20997@cs.brown.edu> mhw@cs.brown.edu (Mark Weaver) writes:
>
>>Typically running X, compiling, and perhaps dragging around xterms
>>or reading manual pages at the same time.  To be honest, I'm not
>>exactly sure that it's thrashing, but I get one hell of a lot of
>>disk activity.  I always get significant disk activity from compiling,
>>even though my /tmp is a ramdisk.
>
>>On Linux, while doing the same thing, there is hardly any disk
>>activity at all.  I always think something's wrong when compiling
>>under Linux because I don't hear the characteristic disk chattering
>>that I associate with compiling.
>
>Actually, what you're experiencing might not be VM swapage at all, but
>rather synchronous writes to disk.  From what I understand, from other
>posts on this subject, Linux delays writing everything to disk until
>buffers get flushed.  Whereas *BSD always writes superblock/inode/
>whatever data immediately to disk, and only delays writing actual data
>blocks until buffers flush.  This causes *BSD to be a little slower in
>asynchronous disk benchmarks and such, but causes it to trash the
>drive much less severely if your machine were to crash, say, in the
>middle of a large compile.

I'm aware of the synchronous writes issue, but again, my /tmp is
a ramdisk, and gcc is using /tmp for it's temporary files (I
checked), and it should only be writing one file per source file,
so I shouldn't be getting NEARLY that much disk chattering.

	Mark
--------------------------------------------------------------------
Email: Mark_Weaver@brown.edu           | Brown University
PGP Key: finger mhw@cs.brown.edu       | Dept of Computer Science