Re: [coman] Device Classes?

[Juergen Schoenwaelder <[e-mail]>]

On Mon, Apr 07, 2014 at 11:57:06AM -0400, Paul Duffy wrote:
> On 4/4/2014 9:04 AM, Juergen Schoenwaelder wrote:
> >On Fri, Apr 04, 2014 at 01:45:47PM +0200, Hannes Tschofenig wrote:
> >>Of course, you could replace CoAP entirely with SNMP (as Juergen had
> >>suggested).
> >>
> >I did not suggest to replace CoAP with SNMP. What I said is:
> >
> >1) We have done an SNMP only implementation and we know it works and
> >    what the resource consumption is.
> 
> Please more detail?  SNMP version?  Security? Implemented MIBs, etc?
> 

http://dx.doi.org/10.1109/MCOM.2012.6384464
http://dx.doi.org/10.1109/GreenCom.2012.21
http://dx.doi.org/10.1007/978-3-642-21484-4_13

In short, we implemented all three versions of SNMP including USM
security. We also did our own DTLS implementation but did not find the
time to integrate it into the SNMP code. Anyway, if you do security in
software, this is a big hit - far bigger than the protocol stuff.

Concerning MIBs, we did a bit of the IF-MIB, the LOWPAN-MIB and the
RPL-MIB plus bits and pieced of the ENTITY-MIB and the
ENTITY-SENSOR-MIB (the later we used to expose an energy sensor
attached to a mote).

Fitting all the MIB modules plus SNMPv3 on the modes is getting tough
since RPL eats quite some resources (but things may have improved
meanwhile).

/js

PS: No additional standards were needed to do any of the SNMP stuff. ;-)

-- 
Juergen Schoenwaelder           Jacobs University Bremen gGmbH
Phone: [phone]         Campus Ring 1, 28759 Bremen, Germany
Fax:   [phone]         <http://www.jacobs-university.de/>