Return-Path: linux-nfs-owner@vger.kernel.org Received: from mx1.redhat.com ([209.132.183.28]:44727 "EHLO mx1.redhat.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1750901AbaCHQ40 (ORCPT ); Sat, 8 Mar 2014 11:56:26 -0500 Message-ID: <531B4BB7.9010303@RedHat.com> Date: Sat, 08 Mar 2014 11:56:23 -0500 From: Steve Dickson MIME-Version: 1.0 To: Neil Brown CC: linux-nfs@vger.kernel.org Subject: Re: [nfs-utils RPC-PATCH 0/4] Add options to nfsd etc to avoid needing to write to /proc References: <20140220063616.6548.42556.stgit@notabene.brown> In-Reply-To: <20140220063616.6548.42556.stgit@notabene.brown> Content-Type: text/plain; charset=UTF-8 Sender: linux-nfs-owner@vger.kernel.org List-ID: On 02/20/2014 01:36 AM, Neil Brown wrote: > There are a number of NFS-related setting that currently must be set > by writing to various files under /proc. > This is a bit clumsy, particularly for systemd unit files. > > So this series adds options to a number of commands where relevant. > > The first two (rdma, and nfsv4{grace,lease}time) I am quite comfortable with. > The third (nlm grace time) I think is probably right but if someone can argue > an alternate approach I'm unlikely to resist. > The fourth is .... uhm. You better look yourself. > > Part of me thinks that nlm port numbers should be set in /etc/sysctl.conf (or sysctl.d) > and /etc/modprobe.d should have something like > > install lockd sysctl -p /etc/sysctl.d/lockd > > but last time I tried that it broke "modprobe --show-depends". > Also it is awkward to get setting from /etc/sysconfig/nfs into /etc/sysctl.d/lockd > > Thoughts? I finally got the cycles to take a look at these... My apologies for taking so long... So I went ahead took a look... Clean them up a bit. There were a couple typos and they did not apply cleanly to my tree... While I was doing this I got this gnawing feeling that we probably should have some type of global configuration file where all these command line variables can be set. It would have to be distro friendly meaning the same place in all distros... Maybe something like /etc/nfsclient.conf patterned off the /etc/nfsmount.conf config file?? So I'm thinking it does make sense to have a way to set all these but I'm just not keen on how they are being set. IDK... Maybe I'm over thinking it.. :-) Finally, during my testing the only flags that seem to work was the statd ones: # rpc.nfsd --rdma 8 rpc.nfsd: Unable to request RDMA services: Protocol not supported # rpc.nfsd --grace-time 66 rpc.nfsd: Unable to set /proc/fs/nfsd/nfsv4gracetime: Device or resource busy # rpc.nfsd --lease-time 66 rpc.nfsd: Unable to set /proc/fs/nfsd/nfsv4leasetime: Device or resource busy Is this expected? steved. > > Thanks, > NeilBrown > > > --- > > Neil Brown (4): > nfsd: add -r and --rdma options to request rdma service. > nfsd: alloc nfsv4leasetime and nfsv4gracetime to be set. > nfsd: set nlm grace time to make NFSv4 grace time > statd: add options to set port number of lockd. > > > utils/nfsd/nfsd.c | 52 +++++++++++++++++++++++++++++++++++++++---- > utils/nfsd/nfsd.man | 21 +++++++++++++++++ > utils/nfsd/nfssvc.c | 59 +++++++++++++++++++++++++++++++++++++++++++++++++ > utils/nfsd/nfssvc.h | 2 ++ > utils/statd/statd.c | 57 +++++++++++++++++++++++++++++++++++++++++++++-- > utils/statd/statd.man | 24 ++++++++++++++++++-- > 6 files changed, 204 insertions(+), 11 deletions(-) >