SCTP Discussions

Text archives Help


Re: AW: [SCTP-Discussion] problem when establishing association with -pre11


Chronological Thread 
  • From: Michael Tüxen <Michael.Tuexen AT micmac.franken.de>
  • To: <discussion AT sctp.de>
  • Subject: Re: AW: [SCTP-Discussion] problem when establishing association with -pre11
  • Date: Fri Jun 8 13:35:01 2001
  • List-id: SCTP Discussions <discussion.sctp.de>

Dear all,
see my comments below.
Best regards
Michael

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

On Fri, Jun 8, 2001, Held Martin ICM N MC MI E6
<Martin.Held AT icn.siemens.de>
wrote:

>Hi Andreas,
>
>are you refering to the following paragraph in RFC2960?
>
> IMPLEMENTATION NOTE: In some cases (e.g., when the implementation
> doesn't control the source IP address that is used for
> transmitting), an endpoint might need to include in its INIT or
> INIT ACK all possible IP addresses from which packets to the peer
> could be transmitted.
This describes the problem.
>
>If so, could you please add a comment to the API description that the
>parameter localAddressList of sctp_registerInstance() MUST contain ALL
>addresses which might be used as source address.
>
This is to extreme. You must only make sure that the source address used
in each packet of an association is included in the address list. The trivial
solution is to include all addresses. But there are other case and
applications
where you can fulfill this requirement.

For example a FreeBSD host having one interface en0 can have

ifconfig en0 192.168.1.1 netmask 255.255.255.0
ifconfig en0 192.169.1.1 netmask 255.255.255.0

so now it is possible to run

discard_server -s 192.168.1.1
discard_server -s 192.169.1.1

in a way that client from 192.168.1.0 will connect to the server running
on 192.168.1.1
and clients from 192.169.1.0 will connect to the server running on
192.169.1.1.

This is working with FreeBSD but not with Linux (at least when I tried
it). The solution is
that when a server sends a packet the IP stack include the source address
being in the right
subnet.

This feature will be used for test purposes.

>How about address auto configuration in ipv6 - wouldn't this work at all ?
>
>Regards,
>Martin
>
>(However this still doesn't solve the initial problem)
autoconfiguration and multiple interfaces in (at least) a challenge. At
the last bakeoff
we provided a IPv6 testbed. We found no solution to do this with
autoconfiguration. The problem
was that all participants had at least two interfaces.
>
>
>
>>
>> Hi Andreas,
>>
>> only one question for understanding :
>>
>> What is the reason for the SCTP-ULP to register with local
>> adresses, if all
>> adddresses of an interface are to be specified ??? Before reading your
>> statement I thought that it migth be allowed to specify a
>> subset of the
>> addresses (for any reason).
>>
>>
>> Mit freundlichen Grüßen / Kind regards
>>
>> Johannes Krupp
>>
>> ______________________________________________________________
>>
>> SIEMENS Siemens AG
>>
>> Information and Communication mobile
>> Networks
>>
>> Tel: +49 89 722 57509
>> Johannes Krupp Fax: +49 89 722 57506
>> Mobile Internet E-Mail:
>> Johannes.Krupp AT icn.siemens.de
>> ______________________________________________________________
>>
>>
>>
>>
>> > -----Ursprüngliche Nachricht-----
>> > Von: Andreas Jungmaier
>> > [mailto:ajung AT exp-math.uni-essen.de]
>> > Gesendet: Freitag, 8. Juni 2001 00:39
>> > An:
>> > discussion AT sctp.de;
>> > Held Martin ICM N MC MI E6
>> > Betreff: Re: [SCTP-Discussion] problem when establishing association
>> > with -pre11
>> >
>> >
>> > -----BEGIN PGP SIGNED MESSAGE-----
>> > Hash: SHA1
>> >
>> > Martin, all,
>> >
>> > you must make sure, that you specify all IPv6 adresses, that
>> > are assigned to your
>> > interface....else it will look like OOTB, and be ignored.
>> > What are the exact commands you typed in on the client and
>> > the server ?
>> > What are the respective adress configurations (from ifconfig).
>> > What is the LAN topology you used for testing ?
>> >
>> > Regards, Andreas
>> >
>
>_______________________________________________
>discussion mailing list
>discussion AT sctp.de
>http://www.sctp.de/mailman/listinfo/discussion






Archive powered by MHonArc 2.6.19+.

Top of page