Chronological Thread 

< Chronological >      < Thread >

---

See my comments below.

Best regards
Michael

--
e-mail:
Michael.Tuexen AT micmac.franken.de

On Tue, Apr 3, 2001, Yiwen Jiang
<Yiwen.Jiang AT tic.siemens.ca>
wrote:

>Hi Michael and Andreas,
>
>Thank you very much for your replies... I'm combining my reply into one
>email...
>
>> > I have been fighting to get the above two sets of demo code
>> to work for a
>> > while now, but with limited success..
>> >
>> > Both program runs, but there seems to generate some output
>> that doesn't make
>> > sense to me, especially on the client side. For example,
>> the client seems to
>> > receive SCTP messages it sent out to the server, and turns
>> around and try to
>> > parse the message, but discovers that the destination port
>> number is not
>> > registered, and the packet is dropped...
>> [Tuexen Michael] Hmm, sounds like the client and servers run
>> on the same host. This 'feature' was never fully
>> supported and will
>> be dropped with the next release. We only support then
>> having multiple
>> sctp processes running with disjoint sets of
>> IP-addresses. This needs
>> the forwarding table to set up in tune with the processes.
>Ahh... I kinda guessed that was the issue, and tried to do this on 2
>machines, and got into another problem... I discovered that if I specify the
>server address to be: 127.0.0.1, the programs (any pair) behave
>differently... but if I specify the actual server address to be somewhat
>reasonable, all works! This is why it didn't work on 2 machines for me
>initially.
>
>My question now would be, shouldn't the SCTP library translates the
>127.0.0.1 to its real net address before sending packets to the client and
>vice versa (well, really the client is another running server)??
What is the "real net address"? This is a difficult question, especially
if IPv6 comes into the game ...
Maybe it would be good for the demo applications that if you use no
-s parameters all possible ip addresses would be used. Something which
has nothing to do with the sctp stuff.
>
>> [Tuexen Michael] How to run the server:
>> ./echo_server -s 192.168.1.1 -V
>> For the client use
>> ./echo_client -s 192.168.1.2 -d 192.168.1.1 -n 1 -V
>Is the -n option a mandatory one? what does it really do?
>
The -n number parameter defines how many chunks the echo_client
will initially send. If it is zero, both sides will reflect data, but
there is no data. So to see real traffic, you must use the -n parameter.
>Thanks!
You are welcome
>
>_______________________________________________
>discussion mailing list
>discussion AT sctp.de
>http://www.sctp.de/mailman/listinfo/discussion

---

• RE: [SCTP-Discussion] appropriate usage of echo_server/echo_clien t and discard_server/t erminal demo programs, Yiwen Jiang, 04/03/2001
  • Re(2): [SCTP-Discussion] appropriate usage of echo_server/echo_clien t and discard_server/t erminal demo programs, Michael Tüxen, 04/03/2001

---

Archive powered by MHonArc 2.6.19+.