Future of the IPv6 CPE survey - Your Input Needed
nanog at 85d5b20a518b8f6864949bd940457dc124746ddc.nosense.org
Sun Feb 6 02:07:44 CET 2011
On Sat, 5 Feb 2011 18:53:45 +0100
MarcoH - lists <mch-v6ops at xs4all.nl> wrote:
> On 4 feb 2011, at 19:28, Mathieu Goessens wrote:
> > On 24/01/2011 15:28, Mirjam Kuehne wrote:
> >> Based on new information we received since the last publication, we
> >> updated the IPv6 CPE matrix:
> >> http://labs.ripe.net/Members/mirjam/ipv6-cpe-survey-updated-january-2011
> >> In order to make this information more useful for a large user base, we
> >> are preparing a detailed survey to gather more structural feedback about
> >> the range of equipment that is currently in use. Not only would we like
> >> you to participate in this survey, but we also ask you for help in
> >> identifying the right survey questions. Please find a call for input on
> >> RIPE Labs:
> > Hi,
> > It could be interesting to re-use the work of the ipv6ready logo  / ipv6forum .
> > They are doing conformance and interoperability testing using a bench of tests  covering IPv6 core protocol, DHCPv6, IPsec and some other protocols (SNMP, SIP...) .
> And that is exactly where we are going to differ. This is aimed at real people and real environments. The first few iterations we did already show some big differences in what the box says and what people experience is. Something can be fully compliant with every RFC and still have issues or being regarded as unusable.
> I personally expect every box hitting the market being RFC compliant. We are interested in what happens when you come home and plug it in.
My past experience has sometimes been the other way around. It doesn't
comply with an RFC, but "works". That's testament to the robustness of
other RFC compliant implementations that are interacting with it. It is
only when you consider that while one device misbehaving is tolerable,
it'd be a nightmare when you have 1000 or more devices doing the same
thing at the same time (and you're on call when it happens e.g. after
a power black out). After that, you start to care very much about a
minimum of baseline level of compliance with RFCs as well as it working
when it is plugged in at home. You also need to care about both that it
does and then how well it recovers from common types of failures e.g.
loss of and recovery from ADSL line sync.
More information about the ipv6-ops