Re: [tcpm] 793bis: TCP Quiet Time

I am under the impression that web servers do something different:
Normally the client sends the FIN, and the server holds FIN wait, but often
discards it before 2MSL to reduce memory pressure, however if the server
receives a new SYN for the same connection 4tuple (before it is discarded),
the server has the option to continue in the same sequence space, one byte
after the FIN.

I haven't looked, but what happens if a connection is in FIN wait, and a
new SYN arrives?

Thanks,
--MM--
The best way to predict the future is to create it.  - Alan Kay

We must not tolerate intolerance;
       however our response must be carefully measured:
            too strong would be hypocritical and risks spiraling out of
control;
            too weak risks being mistaken for tacit approval.

On Wed, Dec 18, 2019 at 8:16 PM Neal Cardwell <ncardwell=
40google.com@dmarc.ietf.org> wrote:

> On Wed, Dec 18, 2019 at 8:28 PM Yuchung Cheng
> <ycheng=40google.com@dmarc.ietf.org> wrote:
> >
> > On Wed, Dec 18, 2019 at 1:33 PM Wesley Eddy <wes@mti-systems.com> wrote:
> > >
> > > I don't think I noticed anyone responding to Gorry's comment below, and
> > > I haven't made any alterations in the 793bis draft with regard to this
> > > (other than fixing some spelling mistakes).  I wanted to pull this into
> > > its own thread in case other people have thoughts or would like to
> > > discuss further what the quiet time concept's relevance is in 2020.
> > I probably am missing something. What's the issue of the text? Linux
> > to my knowledge does
> > not implement this quiet time.
> > But there are other injection attacks or basic checksum issues etc to
> > corrupt TCP. (Serious) application that solely relying on TCP's
> > integrity is signing up for troubles already ...
>
> I would agree with Yuchung's remarks and the general point that the
> "TCP Quiet Time Concept" from RFC 793 does not seem to be of practical
> importance today, from a number of perspectives:
>
> o I'm not aware of any major TCP implementation that actually follows
> the "TCP Quiet Time" approach.
>
> o Most production applications AFAIK end up using SO_REUSEADDR,
> indicating a need to reuse ports quickly and a lack of concern about
> this kind of issue with old packets ending up mixed into new
> connections.
>
> o Applications that care about this should be using cryptography.
>
> o Users today expect to be able to reboot a machine and get back on
> the network in less than 2 minutes.
>
> best,
> neal
>
> _______________________________________________
> tcpm mailing list
> tcpm@ietf.org
> https://www.ietf.org/mailman/listinfo/tcpm
>