[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[6bone] RFC2772 rewrite
On Wed, Nov 13, 2002 at 08:03:52PM +0000, Robert Kiessling wrote:
> Paul Aitken <[email protected]> writes:
> > Presumably if you take your girlfriend away for the weekend then
> > either you persuade her to let you take your laptop computer with you,
> > or you get Mike Cheney to stand in for you, right?
> You have a very good point pointing to a serious conflict.
> You and others operate pTLAs and provide many valuable services to the
> community. However, it's operated as a spare-time activity without,
> for example, guaranteed response times.
> This leads to serious operational impact on the whole IPv6 world. I
> just want to recall the AS1654 incidence, where a hobby pTLA brought
> down significant parts of the global IPv6 network and we were lucky
> enough that the IPv4 upstream was available to turn off the tunnel
> As a result the IPv6 network quality is considerably worse than IPv4,
> and understandably people are reluctant to trust important services to
> I see only two solutions:
> 1. Isolate 6bone and similarly operated one-host-wildly-tunneled sTLAs
> from a production-quality IPv6 core, and widely implement filtering.
> 2. Assure that pTLAs provide a minimum of service.
That's not to say that hobby ISPs can't have good response times though.
Everyone has to start somewhere, and often it's not practical
(predominantly for financial reasons :) to devote 100% of your time (and
of your colleagues' time) to a project. Many IPv4 ISPs have been started
as 'back bedroom' ISPs, and many small-medium providers are still run as
I'd hate to see requirements appertaining to the necessity for 'full
time staff', as in many cases this isn't possible. I think what's
important is that staff are contactable 24/7, and that there will always
be staff around to handle any network issues that may arise.
At the moment the v6 project I'm a part of is only being run part time,
and not for profit. That's not to say it's any less worthwhile than a
full-blown commercial project (I'm finding increasingly that I'm
spending an equal amount of time on my daytime paid job and my evenings
unpaid job), but I can see where people could have a problem with this.
While I'm always careful to ensure that there is always a point of
contact for the v6 services, I do agree that unless a service has
dedicated staff it can be difficult to devote time to fixing any
problems that arise.
Just because a project is only part-time for its staff doesn't mean that
they can't guarantee a response time. I think the trick here is going to
be to ensure that a minimum level of service can be achieved. How this
is determined I have no idea, but I would be the first person to object
if worthwhile projects were turned down the opportunity to take the
first step towards being a full-blown enterprise because of a
requirement that hadn't been fully thought out.
Just my 2 cents :)