From: =?iso-8859-1?Q?Brasseur_Val=E9ry?= Subject: RE: nfs unlink perf ! Date: Tue, 13 Aug 2002 17:05:25 +0200 Sender: nfs-admin@lists.sourceforge.net Message-ID: <8D7D56C6ED3DD411998D009027DC685E0874E222@srv-grp-s1.segin.com> Mime-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Return-path: Received: from hermes4.atos-group.com ([160.92.18.11]) by usw-sf-list1.sourceforge.net with esmtp (Exim 3.31-VA-mm2 #1 (Debian)) id 17edEl-000819-00 for ; Tue, 13 Aug 2002 08:05:35 -0700 To: nfs@lists.sourceforge.net Errors-To: nfs-admin@lists.sourceforge.net List-Help: List-Post: List-Subscribe: , List-Id: Discussion of NFS under Linux development, interoperability, and testing. List-Unsubscribe: , List-Archive: I have check my code, the unlink is done after the close. > -----Original Message----- > From: Trond Myklebust [mailto:trond.myklebust@fys.uio.no] > Sent: Monday, August 12, 2002 8:55 PM > To: Brasseur Val=E9ry > Cc: nfs@lists.sourceforge.net > Subject: Re: [NFS] nfs unlink perf ! >=20 >=20 > >>>>> " " =3D=3D Brasseur Val=E9ry = writes: >=20 > > I am testing an application which to a lot of link/unlink > > server is a NetApp filer, client is linux 2.4.19+nfs patches. >=20 > > under load it seems that the unlink call is very long when all > > over calls response time seems to be OK. is there any = problems > > with the unlink code ? >=20 > There should be no problems, but it depends on what type of test you > are doing... >=20 > The NFS protocol has to work around some nasty compatibility problems > w.r.t. the POSIX expected behaviour of unlink. In particular, POSIX > allows you to do >=20 > open(file); > write(file); > unlink(file); > read(file); > close(file); >=20 > i.e. to unlink an open file... >=20 > In order for the above to work with NFS, one can't immediately unlink > the file when it is open(). Instead, NFS does a 'sillyrename': it > renames the file to something of the form .nfsXXXXX (where XXXX is > some number) and then performs the unlink when the 'close()' has been > performed. >=20 > Would this suffice to explain your 'bad performance'? >=20 > Cheers, > Trond >=20 ------------------------------------------------------- This sf.net email is sponsored by: Dice - The leading online job board for high-tech professionals. Search and apply for tech jobs today! http://seeker.dice.com/seeker.epl?rel_code=31 _______________________________________________ NFS maillist - NFS@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/nfs