[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: netconf and working "on the box"

On Thursday, Mar 13, 2003, at 20:32 America/Montreal, Faye Ly wrote:
To continue on this topic, I don't think any network engineer with a
green field deployment scenario will purposely design a 'single point of
failure' management network. This is more of retro fit into a legacy
network, let's say DSL or APON. Either case doesn't really have any OOB
management except driving all the way to the curb and plug the laptop
into the console port.
In DSL, one can connect a POTS line and modem to the console port.
Lots of folks do that today.

Unclear to me what APON means in this context. Sounds like a telephony
term. Telephony networks are NOT the IETF's focus -- IP networks are.

We are not focused on managing the layer-1 transmission system, but instead
(since this is IETF) are trying to manage things like switches and router
(and access points for wireless networks, since they are like a switch
or router). It might happen to be the case that one could use XMLconf
to manage a layer-1 transmission system, which would be nice, but it
is not the primary focus of this effort (AFAIK).

I guess we can argue in this case what is the
point of using XML then because they also have the legacy OSS to deal
with?
That certainly is a good question for any prospective user to be asking.
Any new technology will need to provide sufficient user/operator value-add
to be worth deploying -- or it won't be widely deployed. I'd like
to see more user/operator (of IP networks) involvement on this list
to get that input.

  But shouldn't we say that out front that we are not addressing
the network management scenario where the in-band is the only means?

We could also say that we are not addressing the network management
scenario where bamboo telegraph is the last mile.  There are many
things we could say.  The current draft does not (to me) appear
to be claiming to be a panacea, nor did it appear (to me) to be
claiming that it should be the only way to manage a device.

As noted before, I'd never rely on any form of in-band management
in any network that I'd be accountable/responsible for.

Where do you propose to draw the line on what to disclaim ?
Also the problem with SNMP, XML over a bunch of stuff plus FTP/TFTP will
probably not do well with a low bandwidth management interface.
Not clear what you mean by that sentence.  Please restate it.


Has anybody look into the wireless management yet?

Yes.  Serial console port to telephone modem to POTS line works fine
for OOB to a wireless access point's console.

Ran
rja@extremenetworks.com


--
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/>