Chronological Thread 

< Chronological >      < Thread >

---

Dear all,

please note that there are two values:
- The one SCTP uses to announce the receiver window.
- The size of the 'real' receive buffer, which is the
receive buffer size for raw IP packets managed at system
level.

On BSD we read that system variable and use half of that. To increase
the receive buffer on a BSD system you can use as root in a shell
sysctl -w net.inet.raw.recvspace=40000
or so. If you start the sctplib after that it makes use of the larger
buffer.

If you just modify the value announced by SCTP and that is larger than
the raw IP receive buffer you will experience heavy packet loss, of course.
This will result in worse SCTP performance.

Best regards
Michael

On Thursday, Dec 5, 2002, at 11:53 Europe/Berlin, Andreas Jungmaier wrote:

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

On Wednesday 04 December 2002 10:57, Oon Heok Kee wrote:

Hi!

Does anyone have any idea what's the default receiver buffer size?

See the file adaptation.c, function adl_open_sctp_socket(),
around lines 674. It is read by a getsockopt() function.
YMMV.

Is there any way to change it (even modifying the stack itself is fine
with me)? If there is, what do I need to change?

See the file adaptation.c, function adl_set_receiveBufferSize(),
around lines 403. It uses a setsockopt() function.
YMMV (as it depends on the OS used, and maybe other things).

I'm trying to stream video using SCTP but I see that the default
receiver buffer 32768 bytes, which is far too small, so I need to increase
it.

Since that only changes the Raw IP Layer socket buffers, you also
need to modify the file distribution.c, around line 1784, where it
says:
sctpInstance->default_myRwnd = myRWND/2;
and set it there to a useful value.

Please note however, that we found out that with a too large buffer
announced and a high speed network, the (RAW IP socket) buffer will
overflow, and you will see strange packet losses.

Hope that helps,

Andreas

P.S.: Please let me know of the results of your tests as that is
very interesting for me to know.
- --
- ----------------------------------------------------------------------- ----
Dipl.-Ing. Andreas Jungmaier | Stop sending E-POSTCARDS !
Computer Networking Technology Group | Send E-MAILS -- Use PGP !
University of Essen |
http://www.exp-math.uni-essen.de/~ajung | http://www.gnupg.org
My PGP Key-ID: D3824AC0 | http://wwwkeys.pgp.net
- ----------------------------------------------------------------------- ----
My PGP Key-Fingerprint: 228C 7C2C 3381 2DD4 9998 2812 5C0B 0B04 D382 4AC0
- ----------------------------------------------------------------------- ----

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (GNU/Linux)

iD8DBQE97zA2XAsLBNOCSsARAqFFAJ9QI2p/PTVe2znTg51+m140TiJbTACfeskh
qU8gLa+GDAgpl4LYA4o0hPY=
=etc3
-----END PGP SIGNATURE-----

_______________________________________________
discussion mailing list
discussion AT sctp.de
http://www.sctp.de/mailman/listinfo/discussion

---

• [SCTP-Discussion] Buffer size, Oon Heok Kee, 12/04/2002
  • Re: [SCTP-Discussion] Buffer size, Andreas Jungmaier, 12/04/2002
    • Re: [SCTP-Discussion] Buffer size, Michael Tüxen, 12/04/2002
  • <Possible follow-up(s)>
  • Re: [SCTP-Discussion] Buffer size, Oon Heok Kee, 12/05/2002
    • Re: [SCTP-Discussion] Buffer size, Michael Tüxen, 12/05/2002
  • Re: [SCTP-Discussion] Buffer size, Oon Heok Kee, 12/05/2002
    • Re: [SCTP-Discussion] Buffer size, Michael Tüxen, 12/05/2002
  • Re: [SCTP-Discussion] Buffer size, Oon Heok Kee, 12/08/2002

---

Archive powered by MHonArc 2.6.19+.