[6bone] routing problem

[Jeroen Massar]

Wim Biemolt wrote:

> ==> From: Gert Doering
> 
> > To clarify this:  this is not an effect of "caching" of any sort.
It's
> > just plain buggy router implementations that don't handle
withdrawals
> > correctly.
> 
> So I can only hope that a buggy router implementation also 
> means a memory
> leak or something like that causing them to reboot in the 
> real near future?
> 
> The only other option I see is switch back to our /35. We are 
> using IPv6
> as production. Being without any decent IPv6 connectivity for too long
> won't do our case for IPv6 any good I'm afraid.
> 
> -Wim -/- SURFnet

There is one good way: find out the AS's who are 'bad' and depeer them
if they don't respond in x time.
This is just the same as for IPv4...

PS: Gert.. did you put an extra page in your presentation saying TLA's
are moving to /32's?
PS2: Gert, don't get a sunburn in Rhodos ;)

Greets,
 Jeroen