ietf
[Top] [All Lists]

Re: Local Cloud Node

2014-10-03 13:30:21


On 03/10/14 19:12, John C Klensin wrote:


--On Friday, October 03, 2014 09:11 -0400 Phillip Hallam-Baker
<phill(_at_)hallambaker(_dot_)com> wrote:

As someone with far too many IP addressable devices in my
house, I am getting rather fed up of every new device
demanding to be hooked up to a service in the cloud.

Yes, I get the fact that the providers are trying to razor and
blade me with their business model. But that model only works
for Gillette because the blade is the most expensive,
difficult part to make.

As a consumer, these services in the cloud represent a lot of
risk and a painful integration nightmare. The point of
installing network addressable switches to control the lights
is so that they can be turned on and off automatically, not so
that I can control them from an iPhone where I have to press
six buttons and wait a minute for the app to load. Then
another quarter hour while it insists on updating itself.

There is therefore a real need for a standards based device
that provides the same services these devices expect from 'the
cloud' and make them available in the local net. And it should
be really easy to connect a newly purchased device to such a
cloud.

While I agree that there is a problem, and largely agree with
your analysis of it, I don't see work here for which the IETF
has the right expertise, at least unless there were some chance
of convincing a WG like Homenet that their boundary router specs
should be expanded adequately to act as a controller for these
families of devices. 

As a more general observation, I think there are a number of
areas where performance and efficiency as seen by end users, as
well as security and privacy, would be well-served by moving
services from shared centralized facilities (e.g., "the cloud")
[back] onto more local networks that fell into the user's
administrative domain and control.    The class of devices
described above are, IMO, just one set of instances of the more
general issue.

(No hats and all that.)

I agree. (And I'm also doing a little work on a side-project
along these lines - wonder how many of us are at stuff like
that:-)

And fwiw, I've not yet hit places so far where there's obvious
IETF work to be done in the short term. Longer term there could
be but I think it'd depend on some of these activities getting
popular first.

Cheers,
S.



    john

Disclaimer: I've been doing some "day job" work that is
consistent with some of the ideas suggested above and that
assumes significantly more powerful edge devices for relatively
small networks than we have gotten used to.  To the best of my
knowledge, there is no plan to bring any aspect of that work to
the IETF so this is not an IPR disclosure, merely an observation
about external discussions that may be influencing my thinking
and comments.




<Prev in Thread] Current Thread [Next in Thread>