linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Luis R. Rodriguez" <mcgrof@kernel.org>
To: NeilBrown <neilb@suse.com>
Cc: "Luis R. Rodriguez" <mcgrof@kernel.org>,
	"Darrick J. Wong" <darrick.wong@oracle.com>,
	Hou Tao <houtao1@huawei.com>,
	linux-xfs@vger.kernel.org, david@fromorbit.com,
	linux-fsdevel@vger.kernel.org, Jeff Mahoney <jeffm@suse.com>,
	Tso Ted <tytso@mit.edu>, Dmitri Pal <dpal@redhat.com>
Subject: Re: Filesystem configuration parsers - (was: Re: [RESEND][PATCH] xfs: add online uevent for mount operation)
Date: Fri, 25 Aug 2017 00:39:03 +0200	[thread overview]
Message-ID: <20170824223903.GV27873@wotan.suse.de> (raw)
In-Reply-To: <87a82ok45i.fsf@notabene.neil.brown.name>

On Fri, Aug 25, 2017 at 08:11:05AM +1000, NeilBrown wrote:
> On Thu, Aug 24 2017, Luis R. Rodriguez wrote:
> >
> > Just saying, we better damn be sure this is the way we want to go.
> >
> 
> Just to provide contrast, NFS has a config file which can provide
> certain defaults for mounted filesystems.  However it doesn't require a
> udev event to achieve this.
> 
> Instead, it provides "/sbin/mount.nfs" which imposes the defaults at
> mount time, rather than just after mount time.
> 
> There could, conceivably, be an "/sbin/mount.xfs" which read the config
> file and set all the defaults concurrently with the mount operation.
> 
> I'm not necessarily saying this is a better way to go, but I agree with
> the need to be sure, and this provides evidence that other approaches
> are possible and worth considering.

I like the NFS approach given:

  o It would mean you can keep things contained within each filesystem
    userspace tool. This would not escalate the requirements for what
    parser to use, ie we would not need to pick one for a higher level
    language for udev rules. Tools which already have a parser
    like e2fsprogs can keep on chugging with libprofile, meanwhile XFS and
    others can choose whatever library their own heart pleases and the
    requirement still remains just simple file parsing, not needing
    higher level language support.

 o One advantage of having things like /sbin/mount.<filesystem> is
   shared filesystems between a type of system can be kept separately
   from host specific mount points. This could allow sharing *one* file
   for a type of target system (database, development, etc). There
   would not be any need to look for your own filesystem in /etc/fstab,
   or adding entries, then /etc/fstab would just become very host specific.

 o Custom entries can be supported without having to worry about
   breaking parsing of the general /etc/fstab, it also removes clutter
   from complex lines and entries out of /etc/fstab

Are there other needs for uevents which could not be addressed through
a filesystem configuration file or which prove more efficient or provide
more flexibility?

  Luis

  reply	other threads:[~2017-08-24 22:39 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
     [not found] <1503582115-37220-1-git-send-email-houtao1@huawei.com>
     [not found] ` <20170824171028.GF4796@magnolia>
2017-08-24 18:04   ` Filesystem configuration parsers - (was: Re: [RESEND][PATCH] xfs: add online uevent for mount operation) Luis R. Rodriguez
2017-08-24 18:13     ` Luis R. Rodriguez
2017-08-24 18:59     ` Luis R. Rodriguez
2017-08-24 22:11     ` NeilBrown
2017-08-24 22:39       ` Luis R. Rodriguez [this message]
2017-08-24 23:34       ` Theodore Ts'o
2017-08-25  0:02         ` Luis R. Rodriguez
2017-08-25  4:54         ` Darrick J. Wong
2017-08-25  1:01     ` Dave Chinner
2017-08-25  1:20       ` Luis R. Rodriguez
2017-08-25  2:18         ` Dave Chinner
2017-08-25  5:29           ` Darrick J. Wong

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170824223903.GV27873@wotan.suse.de \
    --to=mcgrof@kernel.org \
    --cc=darrick.wong@oracle.com \
    --cc=david@fromorbit.com \
    --cc=dpal@redhat.com \
    --cc=houtao1@huawei.com \
    --cc=jeffm@suse.com \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-xfs@vger.kernel.org \
    --cc=neilb@suse.com \
    --cc=tytso@mit.edu \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).