RE: netconf WG charter proposal

["Allen, Keith" <kallen@tri.sbc.com>]

Allen> Or will I have to maintain some type of mapping between the
Allen> identifiers used by XML and those used by SNMP?)

> To tell the truth: I believe that such mappings will at the end be
> unavoidable. Avoiding such mappings means to carry SNMP/SMI naming
> restrictions forward which the operators very well dislike. Given the
> fact that operators are crucial to the success of a new protocol, we
> should think twice whether we not better go ahead and address their
> needs and overcome some ugliness of the past.


Juergen,

I think you are right.  Well, what's one more mapping to keep track of,
anyway?


Keith Allen
SBC Technology Resources
9505 Arboretum Blvd.
Austin, TX 78759
(512) 372-5741
kallen@tri.sbc.com
 

-----Original Message-----
From: Juergen Schoenwaelder [mailto:schoenw@ibr.cs.tu-bs.de] 
Sent: Thursday, April 10, 2003 1:20 PM
To: Allen, Keith
Cc: xmlconf@ops.ietf.org
Subject: Re: netconf WG charter proposal


>>>>> Allen, Keith writes:

Allen> We would prefer to use one protocol for both configuration and
Allen> monitoring both to limit the number of interfaces we have to
Allen> support and to eliminate the problems that crop up with trying
Allen> to use multiple protocols to manage one box.  (Will the SNMP
Allen> trap that notifies me that a port has a problem identify that
Allen> port the same way that I did when I used XML to configure it?

To tell the truth: I believe that only one protocol will survive in
the long term. If the IETF manages to create a new XML-based protocol
which the operators like so much that they push their vendors to
provide solid interoperable implementations, then SNMP is just legacy
and might disappear from operator networks faster than some people
might think.

There is, however, also a serious chance that this new attempt to
solve the management protocol problems we have will fail like many
other attempts to evolve/augment/replace SNMP have failed in the
past. The creation of a working group is just the beginning of a long
and difficult process - and there is a long way to go.

The key to success here is in my opinion strong operator involvement
during the whole process (plus the usual things such as free good
implementations of the protocol as well as software components that
utilize the protocol to manage networks - not elements).

Coming back to your question:

Allen> Or will I have to maintain some type of mapping between the
Allen> identifiers used by XML and those used by SNMP?)

To tell the truth: I believe that such mappings will at the end be
unavoidable. Avoiding such mappings means to carry SNMP/SMI naming
restrictions forward which the operators very well dislike. Given the
fact that operators are crucial to the success of a new protocol, we
should think twice whether we not better go ahead and address their
needs and overcome some ugliness of the past.

[This is just my personal opinion - nobody has to agree with this.]

/js

-- 
Juergen Schoenwaelder    <http://www.iu-bremen.de/>

--
to unsubscribe send a message to xmlconf-request@ops.ietf.org with
the word 'unsubscribe' in a single line as the message text body.
archive: <http://ops.ietf.org/lists/xmlconf/>