All of lore.kernel.org
 help / color / mirror / Atom feed
From: Pablo Neira Ayuso <pablo@netfilter.org>
To: Yury Batrakov <batrakov@gmail.com>
Cc: netfilter@vger.kernel.org
Subject: Re: Question about conntrack
Date: Thu, 18 Sep 2008 16:05:22 +0200	[thread overview]
Message-ID: <48D26022.7060207@netfilter.org> (raw)
In-Reply-To: <786c7f0809050531s1a12b40au8c7b2c9387f9b055@mail.gmail.com>

Yury Batrakov wrote:
> Hello all!
> 
> I've got a couple of questiona about netfilter's connection tracker,
> could someone clarify it to me?
> 1. When conntrack is being flushed? In /proc/net/ip_conntrack I see
> lots of UNREPLIED connections, I reload conntrack kernel module but
> see the table being filled with old entries. The same looks to happen
> after rebooting Linux box.
> 2. Are UNREPLIED connections being wiped when number of connections to
> track equals to conntrack's capacity? Some web resources tell they
> are, but some tell otherwise. I tried to reduce conntrack's capacity
> and saw that these connections aren't wiped and cause conntrack to
> overflow is it bug or feature?

No, when the table gets full the selected conntracks are those that are
!ASSURED.

> 3. I played with NOTRACK target of table raw and discovered that if I
> add a NOTRACK rule that matches with already established connections,
> they stuck in table as unreplied. Most of them disappear when I set
> net.ipv4.netfilter.ip_conntrack_tcp_loose to 0. Is it recommended to
> kill existing unreplied connections in this way?

You may kill the entries using:

# conntrack -D -s IP -p tcp --dport xyz

See conntrack(8) for reference, or the conntrack-tools website.

> Could it be any side
> effect for new or currently established connections that don't match
> NOTRACK?

No, if you really only kill the conntracks that you don't need anymore.

-- 
"Los honestos son inadaptados sociales" -- Les Luthiers

  reply	other threads:[~2008-09-18 14:05 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2008-09-05 12:31 Question about conntrack Yury Batrakov
2008-09-18 14:05 ` Pablo Neira Ayuso [this message]
  -- strict thread matches above, loose matches on Subject: below --
2009-10-24 12:51 Michele Petrazzo - Unipex
2009-10-26 15:52 ` Gary Smith
2004-08-03  0:19 question " Jason Opperisano
2004-08-02 20:32 Paulo Ricardo

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=48D26022.7060207@netfilter.org \
    --to=pablo@netfilter.org \
    --cc=batrakov@gmail.com \
    --cc=netfilter@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.