All of lore.kernel.org
 help / color / mirror / Atom feed
From: Willy Tarreau <willy@w.ods.org>
To: Yoss <bartek@milc.com.pl>
Cc: linux-kernel@vger.kernel.org
Subject: Re: Memory leak in 2.4.33-pre1?
Date: Wed, 15 Feb 2006 21:11:43 +0100	[thread overview]
Message-ID: <20060215201143.GL11380@w.ods.org> (raw)
In-Reply-To: <20060215090130.GA3343@milc.com.pl>

Hi !
On Wed, Feb 15, 2006 at 10:01:30AM +0100, Yoss wrote:
 								 
> > Have you noticed the difference ? So the memory is not wasted at all. It's
> > just reported as 'used'.
> 
> I see. I also noticed that I simply cannot tell what for is this memory
> used. Is this better for me to enlarge cache_mem in squid for about
> 100MB and have less *_cache or is better to have more *_cache? :)

If squid is the main usage of your server, then I guess it should be better
to reserve some memory for it by increasing its cache_mem instead of seeing
this memory wasted as cache for useless files.

> > > > If you don't believe me, simply allocate 1 GB in a process, then free it.
> > > If that what you said is rigth, day after tomorow I'll have the same
> > > situation - only thing I have changed is kernel. So we'll see. :)
> > 
> > If you encounter it, simply run the tool below with a size in kB. Warning!
> > a wrong parameter associated with improper ulimit will hang your system !
> > Ask it to allocate what you *know* you can free (eg: the swapfree space).
> 
> I don't matter is this memory used for cache or free. I just want to be
> sure that it is not leaking :)

I don't remember when "standard" 2.4 was last seen leaking. By "standard",
I mean without fancy drivers and filesystems. For instance, look at my
outgoing dns+mail+http relay (pentium 133 + 96 MB RAM) :

ns# uptime
  9:05pm  up 466 days, 15:10,  1 user,  load average: 0.04, 0.01, 0.00
          ^^^^^^^^^^^
ns# free
             total       used       free     shared    buffers     cached
Mem:         94180      91820       2360          0      11080       5700
-/+ buffers/cache:      75040      19140
Swap:       524656      27108     497548
ns# uname -rv
2.4.18-wt4 #1 Mon Apr 1 13:57:42 CEST 2002
^^^^^^                                ^^^^
Last built in 2002, nearly 4 years ago. After 466 days uptime, I suspect
I would have noticed it if there were such important leaking ;-)

> Bart?omiej Butyn aka Yoss
> Nie ma tego z?ego co by na gorsze nie wysz?o.

Cheers,
Willy


      reply	other threads:[~2006-02-15 20:12 UTC|newest]

Thread overview: 9+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-02-13 21:46 Memory leak in 2.4.33-pre1? Yoss
2006-02-14  0:05 ` Willy Tarreau
2006-02-14  0:34   ` Roberto Nibali
2006-02-14  4:50     ` Willy Tarreau
2006-02-14  8:22     ` Yoss
2006-02-14  8:21   ` Yoss
2006-02-14 21:43     ` Willy TARREAU
2006-02-15  9:01       ` Yoss
2006-02-15 20:11         ` Willy Tarreau [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=20060215201143.GL11380@w.ods.org \
    --to=willy@w.ods.org \
    --cc=bartek@milc.com.pl \
    --cc=linux-kernel@vger.kernel.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 an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.