[bess] BESS charter [was: Re: John Scudder's Discuss on draft-ietf-bess-bgp-sdwan-usage-20: (with DISCUSS and COMMENT)]

[John Scudder <[e-mail]>]

Hi Linda,

To be clear, I hope the BESS chairs and AD will also engage on this question, and I’ve changed the subject line to distinguish this subthread.

> On Feb 27, 2024, at 10:47 PM, Linda Dunbar <[e-mail]> wrote:
> 
> ### Is it in charter?
> 
>  Looking at the BESS charter, I don't see how this document fits. If it weren't
> for the preceding question, I probably wouldn't have thought to look. But as it
> stands, I have to ask: why does the WG think this is a chartered work item?
> 
> [Linda] BESS Charter states “ The BGP Enabled Services (BESS) working group is responsible for
> defining, specifying, and extending network services based on BGP.”
> This document is discussing using BGP for controlling a new network service(i.e., SD-WAN).

You found the one sentence in the charter that I identified too. I think taken in context though, it's clear that BESS isn't chartered to define, specify, and extend every conceivable network service using every conceivable forwarding plane just because BGP is the control plane. This is implied by the very next sentence, “In particular, the working group will work on the following services”, followed by a list of specific services and forwarding planes. If every service that used a BGP control plane were automatically in-scope, there wouldn't be a need for the “in particular” enumeration. To further underscore the point, skip to the next paragraph after that list, "The working group may also suggest new services to be supported by BGP and these may be added to the working group charter **subject to rechartering**.” (Emphasis added.) We might also observe that IPSec VPNs were well-known at the time the BESS charter was written, so if the intent was to put them in scope, the opportunity was there.

BESS hasn’t rechartered.
SD-WAN isn’t a BESS milestone. (Most of the other WG documents are also not milestones. :-o)
The draft in question doesn’t “define, specify, or extend” anything, it’s (I guess) a use cases and applicability document.

None of these things is, by itself, necessarily a deal-breaker, but I do think in aggregate, it motivates a need to have this discussion. Again, I don’t expect that the authors should (or even can) carry the main load here.

Thanks,

—John