mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: "Randy.Dunlap" <rddunlap@osdl.org>
To: DervishD <raul@pleyades.net>
Cc: linux-kernel@vger.kernel.org
Subject: Re: Help with /proc/net/tcp fields
Date: Fri, 20 Feb 2004 10:16:04 -0800	[thread overview]
Message-ID: <20040220101604.0686f351.rddunlap@osdl.org> (raw)
In-Reply-To: <20040220105013.GA5606@DervishD>

On Fri, 20 Feb 2004 11:50:13 +0100 DervishD <raul@pleyades.net> wrote:

|     Hi all :)
| 
|     I'm trying to decode what the field in the output of
| /proc/net/tcp means, with little success. Apart from the fact that
| the four last fields have no description (and looking at undocumented
| sources is really a pain a very time-consuming), I have the following
| doubts:
| 
|     - I assume that 'sl' is the socket number, but, what does 'sl'
| stand for?

'sl is just an index.  Maybe it means 'slot'.

|     - What represents the 'st' field?

socket 'state' -- enumerated in include/linux/tcp.h

|     - I suppose that 'tr:tm->when' represents if there is any timer
| on that socket (well, 'tr' is the number of timers), and when will
| expire the nearest one, but I'm not sure.

Looks right at a quick glance.

|     - I suppose that 'timeout' is the time to die when the socket is
| in FIN_WAIT state, but I'm afraid I'm wrong :(

It's the value of 'probes_out', which has a comment (?) of:
/* unanswered 0 window probes */

|     - Does the 'retrnsmt' field show the retransmissions happened in
| this socket?

Looks like a counter that is incremented up to a threshold...
For ACK handling.

|     If anyone can explain me these fields (and the unnamed fields at
| the end) I will be very grateful, and if someone could direct me to a
| site with related information it will help, too.

Last 2 - 7 unnamed fields:

depending on 'state':

case TCP_SEQ_STATE_TIME_WAIT:
	refcnt bucket_pointer (no trailing 5 fields)

case TCP_SEQ_STATE_OPENREQ:
	refcnt open_request_pointer (no trailing 5 fields)

case TCP_SEQ_STATE_LISTENING:
case TCP_SEQ_STATE_ESTABLISHED:
	refcnt socket_pointer rto ato qpp snd_cwnd snd_ssthresh
where:
  rto == retransmit_timeout
  ato == delayed ACK predicted tick
  qpp == (scheduled number of quick acks << 1) | pingpong(interactive)
  snd_cwnd == sending congestion window
  snd_ssthresh == slow start size threshold (-1 if >= 0xFFFF)


I suggest asking such questions on the netdev mailing list
(netdev@oss.sgi.com).

HTH.
--
~Randy

      reply	other threads:[~2004-02-20 18:26 UTC|newest]

Thread overview: 2+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2004-02-20 10:50 DervishD
2004-02-20 18:16 ` Randy.Dunlap [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=20040220101604.0686f351.rddunlap@osdl.org \
    --to=rddunlap@osdl.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=raul@pleyades.net \
    /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®