[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: Proposed Update to Netconf Charter
On Wed, Jun 29, 2005 at 04:33:01PM -0400, Sharon Chisholm wrote:
> "Additional phase 2 work including:
>
> - Requirements and Guidelines for defining Netconf content to enable
> interoperable, high-quality and usable netconf implementations. Requirements
> will be defined around specification language, access control, compliance,
> backwards compatibility, depicting relationships, event specification, and
> application error message specification.
I also would prefer to talk about "data models" rather than "Netconf
content". I would strike out "high-quality and usable" as it does not
really help in charter text. I am not sure what "backwards compatibility"
means - I guess I know what you have in mind but this phrase can be
read in several ways. I rank guidelines on how to write data models
high.
I think access control should not be lumped together with data model
specification guidelines. This is really a separate point and I am not
sure we are ready to agree how that should be done. I do not rank this
very high at the moment and could very well live with postponing this
to the 3rd phase once we start to have operational experience.
> - An initial set of application-level re-usable data types such as IP
> Addresses, MAC addresses, etc. This definition would be compliant to the
> above defined requirements and guidelines for Netconf content.
This is what I believe is really important to get done. My recent
study on the contents of some ~800 MIB modules shows quite clearly
how important some of these core data types have been in the SMI
world. So I would rank this point high.
> - An XML Schema for reporting information about the Netconf system. This
> definition would be compliant to the above defined requirements and
> guidelines for Netconf content.
I would prefer "A concrete data model" over "An XML schema" to reduce
the number of different terms for the same thing down to one. If we
have concensus to use XML schema, then I am also fine with using XML
schema terminology everywhere. I just dislike the mixture of terminology
here. I rank this medium high - I do believe that we need to write
concrete examples to see whether our guidelines work, but I am not
sure netconf itself has something of reasonable complexity to
configure.
> - A netconf protocol specification for asynchronous messaging to enable the
> sending of events. This must preserve the netconf layers."
Again, not totally clear. Are we talking about events related to a
netconf implementation itself or are we talking about the transport
of general notifications (aka syslog messages)? I do not rank this
very high - could live with postponing this to phase 3.
/js
--
Juergen Schoenwaelder International University Bremen
<http://www.eecs.iu-bremen.de/> P.O. Box 750 561, 28725 Bremen, Germany
--
to unsubscribe send a message to netconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/netconf/>