Multiple component OR Proxy ?

Document created by cdnadmin on Jan 24, 2014
Version 1Show Document
  • View in full screen mode
This document was generated from CDN thread

Created by: Paul Bertrand on 12-04-2011 05:01:03 AM
Hi all,
Is there any advice (best solution, documentation, examples, ...) about solution for non energywise capable components ?
{[Energywise network] <----> } <---> [non energywise network]
{[Energywise network] <----> [Endpoint with multiple compoenent]} <----> [non energywise network]
In advance thanks for your answers or ideas

Subject: RE: Multiple component OR Proxy ?
Replied by: John Parello on 12-04-2011 01:35:13 PM
Hi Paul,

That's a model and design issue so there's no right or wrong answer.

That said think of the SDK as a way to get a device associate to a switch (parent) for aggregation in an EW domain. Since an EW domain is considered a collection of devices that represents one unit of energy management (like all the devices in a floor or a building, rack of a data center, etc) it's best to implement the SDK at the finest grain reporting level of a device (as opposed to a gateway to multiple devices)

So a device implementing the SDK should be something you want to "meter" individually. If you use the SDK in a management app (as a gateway) or some aggregated product it may not link up with a domain (it may overlap).

So my advice is to go with the lowest level you can which looks like your second example.

If you have a set of devices or servers in mind feel free to send along a topology etc of what you have to the mailing list energywise-cdn or post here  and we can take a look at it more specifically.


Subject: RE: Multiple component OR Proxy ?
Replied by: Paul Bertrand on 13-04-2011 09:32:03 AM
Hi John,

The main objective of our project is to enable EW on smart-objects (running IPV6 over our PLC media (WPC from Watteco) ). More precisely, as a first step, these smart-objects are kind of powerline plugs that can be individually commanded remotely and able to deliver their consumption. In a first try their won't be more than 10 to 20 devices.
We've already got a gateway (embeded Linux)  that can route from a LAN to our IPV6 PLC network.
We aim to use this gateway to integrate an EW network, by implementing an endpoint in it. We are already working in recompiling the SDK on our "embed Linux (Openembed)".So we thought about the recent possibility (from V1.2.0) of EW endpoint to define "components".

We are asking ourself, which is the best entry point (documentation, limitations, example, ...) to develop this specific Endpoint with multiple comonents ?


Subject: RE: Multiple component OR Proxy ?
Replied by: John Parello on 14-04-2011 02:09:23 PM
Hi Paul,

WIth about 10-20 devices per "gateway" that makes sense then to implment the SDK there. I wouldn't call that a gateway per se more like a controller so that makes sense with those numbers. The SDK allows you to report back for the controller and any subtended child devices.

If the 10-20 endpoints were in the same EW domain as the controller then the SDK in the controller makes sense. If that controller becomes more like a management station (say controlling devices over multiple domains) then it wouldn't make as much sense.

If needed shoot an email to and we can discuss off forum if you like.