* TCP's reaction to soft errors
@ 2004-09-15 14:44 Fernando Gont
2004-10-01 22:13 ` David S. Miller
[not found] ` <4.3.2.7.2.20041003105509.00e33650@pop.gmx.net>
0 siblings, 2 replies; 3+ messages in thread
From: Fernando Gont @ 2004-09-15 14:44 UTC (permalink / raw)
To: netdev
Folks,
I'm the author of an IETF Internet Draft that discusses a problem arising
from the current TCP's reaction to soft errors. The problem has to do with
long delays in connection establishment attempts that may arise in a number
of scenarios.
The draft proposes to change TCP's reaction to soft errors so that
connections that are in the SYN-SENT or SYN-RECEIVED states are aborted
upon receipt of an ICMP error message that indicates a soft error.
The draft can be found at:
http://www.ietf.org/internet-drafts/draft-gont-tcpm-tcp-soft-errors-00.txt .
I've also put online a "Discussion List" I had prepared for the last IETF
meeting, which has a somewhat expanded explanation of the rationale behind
the proposed fix. This document can be found at:
http://www.gont.com.ar/drafts/TCP-Soft-Errors-Discussion-List.doc .
I'd appreciate any comments on the draft. Thus, I'd be able to address your
comments in the next revision of the draft, and will also sum-up your
feedback and post it to the relevant IETF mailing list (that of the TCPM WG
mailing-list).
BTW, I've set up a page at
http://www.gont.com.ar/drafts/tcp-soft-errors.html so that you have all the
documents available from a single place.
--
Fernando Gont
e-mail: fernando@gont.com.ar || fgont@acm.org
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: TCP's reaction to soft errors
2004-09-15 14:44 TCP's reaction to soft errors Fernando Gont
@ 2004-10-01 22:13 ` David S. Miller
[not found] ` <4.3.2.7.2.20041003105509.00e33650@pop.gmx.net>
1 sibling, 0 replies; 3+ messages in thread
From: David S. Miller @ 2004-10-01 22:13 UTC (permalink / raw)
To: Fernando Gont; +Cc: netdev
On Wed, 15 Sep 2004 11:44:14 -0300
Fernando Gont <fernando@gont.com.ar> wrote:
> The draft proposes to change TCP's reaction to soft errors so that
> connections that are in the SYN-SENT or SYN-RECEIVED states are aborted
> upon receipt of an ICMP error message that indicates a soft error.
I have verified that Linux behaves in a way compliant to this
proposal. No changes are necessary.
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: TCP's reaction to soft errors
[not found] ` <4.3.2.7.2.20041003105509.00e33650@pop.gmx.net>
@ 2004-10-03 21:09 ` David S. Miller
0 siblings, 0 replies; 3+ messages in thread
From: David S. Miller @ 2004-10-03 21:09 UTC (permalink / raw)
To: Fernando Gont; +Cc: netdev
On Sun, 03 Oct 2004 10:57:36 -0300
Fernando Gont <fernando@gont.com.ar> wrote:
> At 15:13 01/10/2004 -0700, David S. Miller wrote:
>
> > > The draft proposes to change TCP's reaction to soft errors so that
> > > connections that are in the SYN-SENT or SYN-RECEIVED states are aborted
> > > upon receipt of an ICMP error message that indicates a soft error.
> >
> >I have verified that Linux behaves in a way compliant to this
> >proposal. No changes are necessary.
>
> Great! Thanks for your confirmation.
> BTW, do you have any record of how long Linux has been behaving this way?
> (I'd like to include this information in the draft).
Even 2.0.x kernels behave this way. Others are welcome to
search back even further :-)
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2004-10-03 21:09 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-09-15 14:44 TCP's reaction to soft errors Fernando Gont
2004-10-01 22:13 ` David S. Miller
[not found] ` <4.3.2.7.2.20041003105509.00e33650@pop.gmx.net>
2004-10-03 21:09 ` David S. Miller
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).