Re: [manet] New DLEP extension draft for WG

[Henning Rogge <[e-mail]>]

On Thu, Jan 28, 2021 at 7:55 PM Rex Buddenberg <[e-mail]> wrote:
>
> Henning,
>
> On first read, this appears very sensible. Raises one question in my
> mind: how closely do these listings parallel MIB variables? Is there
> profit in casting these values as MIB extensions which would allow a
> network management console several hops away to do its job?

(I am replying to the mailing list so that all people can see it)

Hi,

I don't see any parallels with MIBs... I am not even aware of any MIB
for radio configuration at the IETF.

During the last years I have talked and worked with a few Radio
vendors and when asking about how to get metric and runtime data the
topic SNMP never came up. If you are lucky you get a REST interface
(and can point them to IETF standards like RESTCONF/RFC8040), if you
are unlucky you get a proprietary protocol on a proprietary network
socket.

There is also the fact that modern radio systems don't use fixed
values of frequency and bandwidth. Wifi 6 access points (as an
example) can use frequency division multiplexing to talk to several
clients at once... LTE (and earlier cellphone networks) have been
using dynamic time slot and bandwidth allocation for ages. This means
that the values can be neighbor specific, they can change quite often
and there might be radio systems where the router can ask for a better
radio band or more bandwidth.

Henning Rogge