From mboxrd@z Thu Jan 1 00:00:00 1970 From: Greg Bradner Subject: Re: nfs optimization settings in the automounter Date: Tue, 18 Nov 2003 11:11:34 -0800 Sender: autofs-bounces@linux.kernel.org Message-ID: <3FBA6EE6.9010205@rhythm.com> References: <3FBA5C63.9080402@scout.wisc.edu> <3FBA67F7.1050505@zytor.com> Mime-Version: 1.0 Content-Type: multipart/mixed; boundary="===============69509215323673335==" Return-path: In-Reply-To: <3FBA67F7.1050505@zytor.com> List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: autofs-bounces@linux.kernel.org To: "H. Peter Anvin" Cc: autofs@linux.kernel.org This is a multi-part message in MIME format. --===============69509215323673335== Content-Type: multipart/alternative; boundary="------------000001060503070900000805" This is a multi-part message in MIME format. --------------000001060503070900000805 Content-Type: text/plain; charset=us-ascii; format=flowed Content-Transfer-Encoding: 7bit It's an nfs limitation. Reconfig your server and change this option: "include/linux/nfsd/const.h" in the kernel source shows the following definition /* * Maximum blocksize supported by daemon currently at 8K */ #define NFSSVC_MAXBLKSIZE 8192 H. Peter Anvin wrote: >Justin Rush wrote: > > >>I am having a hard time telling the automounter to use the options >>rsize=32768,wsize=32768 when automounting home directories. I have >>tried specifying these options in the auto.home file on my nis master >>server ( -rw,soft,rsize=32768,wsize=32768 nismaster:/export/home/jrush ) >>I have also tried to put them in the localoptions setting in the autofs >>init file ( localoptions='rsize=32768,wsize=32768' ) on both the nis >>master and on the client machines. >>I also tried to specify these options in the auto.master file on the nis >>master (the client machines don't have auto.master files because they >>read the nis master's file ) ( /etc/auto.home >>--timeout=43200,rsize=32768,wsize=32768 ) >>I have also tried to use lower settings such as 16384 >> >>however when I look in /proc/mounts on the client machine it is either >>mounted with 4096 or 8192 >> >> >> > >Are you using NFSv3? As far as I know I think 8192 is max for NFSv2, at >least for most of the NFSv2 servers I know. > >Try doing the mounting without using autofs. If the same thing happens, >it's an NFS issue, not an autofs issue. > > -hpa > >_______________________________________________ >autofs mailing list >autofs@linux.kernel.org >http://linux.kernel.org/mailman/listinfo/autofs > > -- ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ Rhythm & Hues 5404 Jandy Place Los Angeles, CA 90066 Voice: 310 448-7763 Fax: 310 448-7600 gregb@rhythm.com --------------000001060503070900000805 Content-Type: text/html; charset=us-ascii Content-Transfer-Encoding: 7bit

It's an nfs limitation. Reconfig your server and change this option: 

"include/linux/nfsd/const.h" in the kernel source shows the following
definition

/*
 * Maximum blocksize supported by daemon currently at 8K
 */
#define NFSSVC_MAXBLKSIZE       8192

H. Peter Anvin wrote:

Justin Rush wrote:

I am having a hard time telling the automounter to use the options
rsize=32768,wsize=32768 when automounting home directories.  I have
tried specifying these options in the auto.home file on my nis master
server ( -rw,soft,rsize=32768,wsize=32768 nismaster:/export/home/jrush )
I have also tried to put them in the localoptions setting in the autofs
init file ( localoptions='rsize=32768,wsize=32768' ) on both the nis
master and on the client machines.
I also tried to specify these options in the auto.master file on the nis
master (the client machines don't have auto.master files because they
read the nis master's file ) ( /etc/auto.home 
--timeout=43200,rsize=32768,wsize=32768 )
I have also tried to use lower settings such as 16384

however when I look in /proc/mounts on the client machine it is either
mounted with 4096 or 8192


Are you using NFSv3?  As far as I know I think 8192 is max for NFSv2, at
least for most of the NFSv2 servers I know.

Try doing the mounting without using autofs.  If the same thing happens,
it's an NFS issue, not an autofs issue.

	-hpa

_______________________________________________
autofs mailing list
autofs@linux.kernel.org
http://linux.kernel.org/mailman/listinfo/autofs

-- 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rhythm & Hues
5404 Jandy Place
Los Angeles, CA 90066
Voice: 310 448-7763
Fax:   310 448-7600
gregb@rhythm.com

--------------000001060503070900000805-- --===============69509215323673335== Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit Content-Disposition: inline _______________________________________________ autofs mailing list autofs@linux.kernel.org http://linux.kernel.org/mailman/listinfo/autofs --===============69509215323673335==--