*BSD News Article 45868


Return to BSD News archive

Path: sserve!newshost.anu.edu.au!harbinger.cc.monash.edu.au!simtel!zombie.ncsc.mil!news.mathworks.com!uunet!in1.uu.net!usc!sdd.hp.com!hpscit.sc.hp.com!hpax!cupnews2.cup.hp.com!raj
From: raj@cup.hp.com (Rick Jones)
Newsgroups: comp.unix.bsd.freebsd.misc
Subject: Re: NFS write performance
Date: 12 Jun 1995 18:13:34 GMT
Organization: Information Networks Division: Hewlett Packard
Lines: 16
Message-ID: <3ri04e$eog@hpindda.cup.hp.com>
References: <802730803PCTIN@lot.cs.kiev.ua> <3rh8tm$mhg@bonnie.tcd-dresden.de>
Reply-To: raj@cup.hp.com
NNTP-Posting-Host: hpindio.cup.hp.com
X-Newsreader: TIN [version 1.2 PL2]

J Wunsch (j@narcisa.sax.de) wrote:
: Certainly not.  First, since 2.0.5R left already the door (and there
: had been a couple of weeks of a code freeze before), second since NFS
: v2 mandates synchronous writes.  What you're looking for is defined in
: the NFS v3 specs.

I'll pick the nit here - the NFS V2 spec mandates "stable storage" but
(to my knowledge) does not specify the implementation mechanism (nor
stability characteristics?) of that storage - for example, I do not
think it explicitly calls for synchronous writes to disc. If the spec
did, then all those folks using NVRAM caches would also be in
violation of the spec. Of course, this leads to a considreable,
periodic thrashing on the topic of "what is stable storage."

rick jones