- From: Thomas Dreibholz <dreibh AT exp-math.uni-essen.de>
- To: Le Vinh Dat <vinhdat82 AT yahoo.com>
- Cc: discussion AT sctp.de
- Subject: Re: [SCTP-Discussion] SCTP performance measure
- Date: Fri Jul 23 09:42:03 2004
- List-id: SCTP Discussions <discussion.sctp.de>
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
Am Donnerstag, 22. Juli 2004 17:13 schrieb Le Vinh Dat:
>
Hi SCTP community,
>
I'm running sctplib-1.0.1 and socketapi-1.3.1 in Redhat 9.0
>
Now I would like to measure the performance of SCTP vs TCP, how to do that?
>
The measure program should work for both of SCTP and TCP.
>
I found out there was a program called iperf, however, this program seems
>
only work for TCP and LKSCTP.
If this program does *not* use fork() to create additional processes, it
should be no really problem to adapt it to sctplib's socket API. However, our
sctplib is a userland implemenation and the socket API relies strongly on
threads. That is, comparing the performance of kernel TCP and userland
sctplib + socket API, the result will be that kernel TCP is much more
efficient (much less CPU load for the same transmitted bandwidth; I assume
that is what you want to examine).
Best regards
- --
=======================================================================
Dipl.-Inform. Thomas Dreibholz
University of Essen, Room ES210
Inst. for Experimental Mathematics Ellernstraße 29
Computer Networking Technology Group D-45326 Essen/Germany
- -----------------------------------------------------------------------
E-Mail:
dreibh AT exp-math.uni-essen.de
Homepage:
http://www.exp-math.uni-essen.de/~dreibh
=======================================================================
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.4 (GNU/Linux)
iD4DBQFBAMEG32BbsHYPLWURApeAAJdkPtnm5osHv7GuTnE5g+DOfCz7AJ99prc9
KRsWc5mdt8HKwqmz7brceQ==
=tIR0
-----END PGP SIGNATURE-----
Archive powered by MHonArc 2.6.19+.