Job Snijders wrote on 08/06/2024 07:29:
On Fri, Jun 07, 2024 at 06:58:26PM -0400, Jeffrey Haas wrote:
The torches and pitchforks operator crowd that rammed through large
communities in the current form weren't interested in slowing down and
discussing how that'd work.
In terms of WKCs, of the 15 currently defined, all of them signal
actions which are independent of the identity of the source or target
asn. There wasn't a lack of interest or a lack of informed discussion at
the time - just an understanding that the concept of "well known" is
independent of the bit size of the asns on either side of a bgp session.
At best "torches and pitchforks operator crowd" is an incrowd joke, at
worst it could be taken to be a disparaging term. But do I recognize
that many people were unhappy how between RFC 4893 (May 2007) and RFC
8092 (February 2017) IDR was unable to publish a solution to fill the
shoes of Classic Communities.
There certainly was frustration and growing alarm, yes. There was no
portable protocol-level mechanism for inter-domain signaling with
asn32s, and this was accompanied by a realisation that it normally takes
years to get new features rolled out in router bgp stacks. All the
while, the asn16 train was barrelling towards the end of the tracks. Ops
engineers are in the unfortunate position that they have to make things
work, and if the strut design hasn't made it off the blackboard, that
makes it difficult to keep the trucks on the road.
Large communities remind me of the "what the customer really wanted"
cartoon: they're minimalistically simple, and do what's needed for most
inter-domain signaling needs. They will be around for a long time, not
because they're all we have to work with, but because they're a
fundamentally sound engineering mechanism for doing the sorts of things
that operators need to do.
Nick
_______________________________________________
GROW mailing list -- [email protected]
To unsubscribe send an email to [email protected]
