Re: [coman] [6lo] WG Review: IPv6 over Networks of Resource-constrained Nodes (6lo)

[Benoit Claise <[e-mail]>]

On 24/09/2013 15:49, Juergen Schoenwaelder wrote:
> On Tue, Sep 24, 2013 at 03:41:38PM +0200, Benoit Claise wrote:
>> Hi,
>>> Hi,
>>>
>>> I have a couple of issues with the way this charter is worded at this point, which makes the scope of the proposed WG unclear to me on some respects.
>>>
>>> 1. There is careful wording about 'link layer technologies of interest' and about 'Security and management work that is not specific to the link layers being worked on is out of scope' - but no indication what are the criteria to identify these, or a list of initial such layer technologies. Is this completely open by now? Is there a process to discuss and select those?
>>>
>>> 2. 'Selected MIB modules' - Did the team working on this proposal make any preparation work and determine that writing MIB modules is the appropriate way of meeting the manageability requirements for 6lo? If yes, where is this documented? If no, maybe this item should be worded in a more general manner and talk about requirements for managing the IP-over-foo specifications, followed by development of the appropriate information models and data models (the latest may be MIB modules, or may be something else).
>> I would add, for the 6lo people benefits, that
>> http://tools.ietf.org/html/draft-ersue-constrained-mgmt-03 would be
>> a good start.
>> Disclaimer: maybe this draft was mentioned already. I'm not
>> following the 6lo mailing list.
> My understanding is that the charter refers to
>
> http://tools.ietf.org/html/draft-schoenw-6lowpan-mib-03
>
> and that it not aims at a general discussion about network management
> of constrained devices. That is, it refers to concrete work instead of
> abstract work.
That's actually my point. Should we focus on developing a MIB module, 
taking for granting that SNMP is THE way to manage constrained nodes? Or 
should we ask ourselves: based on the collected management requirements, 
let's see what is more appropriate?
You know, like I2RS is currently doing, instead of jumping to 
NETCONF/YANG directly because it seemed like the solution.

Regards, Bneoit

>
> /js
>