By the way, to add some IPv6 technical noise: Have you ever tried to set up MD5 auth. in a BGP over IPv6 session between an Alcatel 7750SR and a Cisco 6500... I keep getting a log message (Cisco) exposing some issue with the MD5 digest from the Alcatel. The funny thing is that Cisco <---> Cisco goes Fine, and Alcatel <---> Alcatel goes fine (as obviously expected)... so may be you know something about any difference in the implementation. Even more "funny" is the fact that BGP MD5 auth. Cisco <---> Alcatel goes perfect over IPv4. I'm also looking at some kind of issue like changes in TCP sequence numbers after MD5 hash or such things, but I found nothing yet... I've talk to Alcatel, but no answer yet... Thx, Nic. michael.dillon@bt.com wrote:
<vendor ears == ON> Hopefully... ;-)
Not likely! This is a motley crew of people who like to jabber, not a forum for your favorite vendor's customer support.
I want to be able to carry IPv6 in a VPRN without having to pay an order of magnitude more for an IOM.
May I suggest that you will make much more impact on your vendor's radar if you explicitly ask your vendor rep for this feature, in writing. And then proceed to try setting up IPv6 in a VPRN as a lab experiment, then raise cases with your vendor's TAC when you run across problems. Tell them that you are doing the lab work as the first phase to commercial introduction of IPv6 services.
One reason that IPv6 support is in a sorry state is that people try something, find it doesn't work, and then suffer in silence. Please do not suffer in silence! Please do explicitly tell people about the specific issues that you encounter and please do name names. Tell us which vendor, which device, which software release, etc. etc. yadda yadda boom!
The more talk there is about IPv6, the more it gets on vendor radar and the quicker they will fix problems and improve support. And don't forget all those vendors of NMS and OSS software.
--Michael Dillon
_______________________________________________ NANOG mailing list NANOG@nanog.org http://mailman.nanog.org/mailman/listinfo/nanog