SCTP Discussions

Text archives Help


Re: [SCTP-Discussion] Single host


Chronological Thread 
  • From: Andreas Jungmaier <ajung AT exp-math.uni-essen.de>
  • To: Michael Tuexen <Michael.Tuexen AT micmac.franken.de>
  • Cc: Alexander Litvin <archer AT whichever.org>, discussion AT sctp.de
  • Subject: Re: [SCTP-Discussion] Single host
  • Date: Fri Jul 16 12:32:02 2004
  • List-id: SCTP Discussions <discussion.sctp.de>
  • Organization: University Duisburg-Essen

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Hi all,

indeed we will try to get around this limitation in a
new sctplib version (supposedly 1.3.0) by dividing the
applications into

1. one sctpd process that only runs once on a host system.
This will require root privileges for opening the raw socket.

2. one or more application processes, that connect to the
sctpd via a local interprocess communication
(unix domain sockets).

This closely reflects the notion that there needs to be
one (!) handler for SCTP communication per system, that can
decide whether there is an association/server or not.
This process can also send ABORTs.

We will release a new version of SCTPLIB shortly that implements
this "separation of powers".

Stay tuned !

Andreas

Am Freitag, 16. Juli 2004 10:00 schrieb Michael Tuexen:
> Hi,
>
> this is known and the workaround is to disable the
> OOTB handling.
>
> The examples provided in sctplib/programs do this
> when the -i option is used.
>
> If you write your own code, do the following:
>
> SCTP_getLibraryParameters(&params);
> params.sendOotbAborts = 0;
> SCTP_setLibraryParameters(&params);
>
> See sctplib/programs/terminal.c for example.
>
> Best regards
> Michael
>
> On Jul 16, 2004, at 3:56 AM, Alexander Litvin wrote:
> > Hi, all!
> >
> > I'm new to sctplib (though not new to SCTP). Just wanted to do
> > some testing. So, I downloaded sctplib-1.0.1 on my FreeBSD 5.1,
> > compiled it, tried to run. And here's my problem.
> >
> > I wrote a simple client and server applications, and tried to
> > run them on the same host. No luck.
> >
> > Doing some GDB, I found that client receives INIT which it
> > sends to server (and, sure, server receives INIT ACK which it
> > sends to client).
> >
> > Apparently, the nature of raw sockets is responsible for this:
> > as long as you just have raw socket on the machine and protocol
> > matches, the socket will receive all the packets for that protocol.
> >
> > Maybe, I'm missing something here. So, before I start hacking around
> > it, could somebody enlighten me? I suppose this "feature" should not
> > be news to the experts here. Is it supposed to be like this? Or is
> > there some "magic" compile/runtime option to overcome this restriction?
> >
> > TIA,
> > Alex
> >
> > _______________________________________________
> > discussion mailing list
> > discussion AT sctp.de
> > http://www.sctp.de/mailman/listinfo/discussion
>
> _______________________________________________
> discussion mailing list
> discussion AT sctp.de
> http://www.sctp.de/mailman/listinfo/discussion

- --
Dipl.-Ing. Andreas Jungmaier
Computer Networking Technology Group
University of Duisburg-Essen
http://www.exp-math.uni-essen.de/~ajung
PGP Key-ID: D382 4AC0

PGP Key-Fingerprint: 228C 7C2C 3381 2DD4 9998 2812 5C0B 0B04 D382 4AC0
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)

iD8DBQFA9620XAsLBNOCSsARAtObAJsHSjm1Ee5D9ISJmkzM6z05Rp04vACeIdTA
5eDFzRC36RYQVca2o0nDMV4=
=dTru
-----END PGP SIGNATURE-----




Archive powered by MHonArc 2.6.19+.

Top of page