linux-fsdevel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Al Viro <viro@ZenIV.linux.org.uk>
To: Alexander Shishkin <virtuoso@slind.org>
Cc: Valdis.Kletnieks@vt.edu, linux-fsdevel@vger.kernel.org,
	akpm@linux-foundation.org, linux-kernel@vger.kernel.org
Subject: Re: [RFC][PATCHv3] List per-process file descriptor consumption when hitting file-max
Date: Wed, 13 Jan 2010 22:57:16 +0000	[thread overview]
Message-ID: <20100113225715.GE19799@ZenIV.linux.org.uk> (raw)
In-Reply-To: <20100113224459.GA21934@ZenIV.linux.org.uk>

On Wed, Jan 13, 2010 at 10:44:59PM +0000, Al Viro wrote:
> On Mon, Jan 11, 2010 at 11:38:07AM +0200, Alexander Shishkin wrote:
> > When a file descriptor limit is hit, display the top consumers of
> > descriptors so that it is possible to identify and fix those which
> > leak them.
> > 
> > Two new sysctl tunables are introduced:
> >   * file-max-consumers -- number of processes to display (defaults
> >     to 10);
> >   * file-max-rate-limit -- time interval between subsequent dumps
> >     (defaults to 10 seconds).
> 
> That *still* doesn't answer the most important question: what for?

BTW, even leaving that (and obvious deadlocks) aside, this stuff is
monumentally bogus.  A process can easily have shitloads of opened
descriptors and very few opened files (see dup() and friends).  Conversely,
you can have shitloads of opened files and not a single opened descriptor
(see mmap()).  And you are calling that when we have too many opened
struct file.

      reply	other threads:[~2010-01-13 22:57 UTC|newest]

Thread overview: 11+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-06-08 11:38 [PATCH] [RFC] List per-process file descriptor consumption when hitting file-max alexander.shishckin
2009-07-29 16:17 ` Alexander Shishkin
2009-07-30 12:40   ` Valdis.Kletnieks
2009-10-11 12:17     ` Alexander Shishkin
2010-01-10 16:34     ` [RFC][PATCHv2] " Alexander Shishkin
2010-01-13 22:12       ` Andrew Morton
2010-01-11  9:38     ` [RFC][PATCHv3] " Alexander Shishkin
2010-01-11 12:40       ` Andreas Dilger
2010-01-13 22:06       ` Andi Kleen
2010-01-13 22:44       ` Al Viro
2010-01-13 22:57         ` Al Viro [this message]

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=20100113225715.GE19799@ZenIV.linux.org.uk \
    --to=viro@zeniv.linux.org.uk \
    --cc=Valdis.Kletnieks@vt.edu \
    --cc=akpm@linux-foundation.org \
    --cc=linux-fsdevel@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=virtuoso@slind.org \
    /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).