mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Sami Kerola <kerolasa@iki.fi>
To: linux-doc@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, davem@davemloft.net,
	Sami Kerola <kerolasa@iki.fi>
Subject: [PATCH] Documentation/networking/ip-sysctl.txt tcp_tw_recycle & tcp_tw_reuse
Date: Mon,  4 Oct 2010 19:32:04 +0200	[thread overview]
Message-ID: <1286213524-4659-1-git-send-email-kerolasa@iki.fi> (raw)

From: Sami Kerola <kerolasa@iki.fi>

Instead of saying 'consult your technical export' the
documentation needs to assist an export to decide whether to keep
these settings off or set them on.

Signed-off-by: Sami Kerola <kerolasa@iki.fi>
---
 Documentation/networking/ip-sysctl.txt |   20 +++++++++++++-------
 1 files changed, 13 insertions(+), 7 deletions(-)

diff --git a/Documentation/networking/ip-sysctl.txt b/Documentation/networking/ip-sysctl.txt
index f350c69..01e2948 100644
--- a/Documentation/networking/ip-sysctl.txt
+++ b/Documentation/networking/ip-sysctl.txt
@@ -444,15 +444,21 @@ tcp_tso_win_divisor - INTEGER
 	Default: 3
 
 tcp_tw_recycle - BOOLEAN
-	Enable fast recycling TIME-WAIT sockets. Default value is 0.
-	It should not be changed without advice/request of technical
-	experts.
+	Enable fast recycling TIME-WAIT sockets. See also caution
+	wrote to tcp_tw_reuse. Default value is 0.
 
 tcp_tw_reuse - BOOLEAN
-	Allow to reuse TIME-WAIT sockets for new connections when it is
-	safe from protocol viewpoint. Default value is 0.
-	It should not be changed without advice/request of technical
-	experts.
+	Allow to reuse TIME-WAIT sockets for new connections when it is safe
+	from protocol viewpoint. One should enable also tcp_timestamps when
+	this is enabled. Default value is 0.
+
+	Caution. Both tcp_tw_recycle and tcp_tw_reuse can cause problems. You
+	should not enable either without being sure about network topology in
+	between node(s) which are using, or used by, the node(s) where the
+	parameter is enabled. Connections that go via TCP state aware nodes,
+	such as firewalls, NAT devices or load balancers may start drop frames
+	because of the setting. The more there are connections the greater is
+	the likelyhood you will see this issue.
 
 tcp_window_scaling - BOOLEAN
 	Enable window scaling as defined in RFC1323.
-- 
1.7.3.1


             reply	other threads:[~2010-10-04 17:32 UTC|newest]

Thread overview: 3+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2010-10-04 17:32 Sami Kerola [this message]
2010-10-05  0:23 ` matt mooney
2010-10-06  9:09 ` Jack Stone

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=1286213524-4659-1-git-send-email-kerolasa@iki.fi \
    --to=kerolasa@iki.fi \
    --cc=davem@davemloft.net \
    --cc=linux-doc@vger.kernel.org \
    --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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®