Re: [Idr] [GROW] BGP Looking Glass Capability

[Robert Raszuk <[e-mail]>]

Hi Rayhaan,


> I guess a good starting point would be to reach out to IDR folks / authors
> of the operational message draft and get their input as to why it didn't
> progress further since that would be useful to guide any revival attempts.
>

Good idea. As I am co-author of this draft taking the liberty to do it
right here :)

A bit of history  - in min 2010 I wrote
https://tools.ietf.org/html/draft-raszuk-bgp-diagnostic-message-00.

Then we spoke about it, trimmed a bit and formed what we considered most
important operationally into
https://tools.ietf.org/html/draft-ietf-idr-operational-message-00

The draft was having good support during the IDR WG adoption and hence it
became WG doc.

Since then most of the authors left the vendor world and our influence to
implement it in significant commercial BGP code bases was no longer
sufficient. Yet vendors told we will implement it if customers ask for it.
So we are 9+ years and still I am not sure if anyone cares much about
switching phone and email channels between NOCs into more programmatic way.

Yet we do see from time to time a pop request to some form to tell peer
ascii strings like sms by BGP, pass some well known address, etc ...

I think this is the fundamental challenge in how we operate peering
relations. I have seen a lot of good automation happening in the IX world,
but when trying to establish BGP sessions today it seems still emails, xls,
word documents style ...

While it does not need to be all TLVs as listed in the
https://tools.ietf.org/html/draft-ietf-idr-operational-message-00 draft but
I think operational message is indeed needed.

Cheers,
R.