Re: [Sidrops] example

Do you still need these examples?

P.S.: Sorry for the late reply.
I don't know how you guys manage to be on so many lists at the same time
and be so active on all of them.

I have a few cases that I believe I can cite.

The most common case of this is small ISPs that, due to improper capacity
planning with their transit providers, or low-budget, end up doing
something that I consider almost criminal.
"Balancing" of incoming traffic based on advertising more specific routes.

I don't think it fits here to detail this practice...
But for them it is quite common to announce a /24 to an upstream to lower
the volume of traffic coming from the other upstream.
Sometimes even as a desperate measure, removing the /22 announcement to the
upstream where 98% of the capacity is being used.

And it is at this moment, if the ROAs are not covering this change in
advertising, that the problem appears.
It doesn't get to "not talking to half the Internet", but the problem does
happen.

And that's one of the reasons we see so many overly permissive ROAs. To get
around this kind of side effect of bad practice.

Yes, I know it's a HORRIBLE practice!
And when interacting with this type of client, our work is more like a
psychologist/psychiatrist than a network operator.
But it really happens.

Em ter., 3 de jan. de 2023 às 13:21, Randy Bush <randy@psg.com> escreveu:

> do we have a real world example where someone, unintentionally i
> presume, published a ROA which invalidated their BGP announcement(s)?
> or changed announcing AS and forgot to issue the new ROA until they
> realized they could not reach half the internet?
>
> randy
>
> _______________________________________________
> Sidrops mailing list
> Sidrops@ietf.org
> https://www.ietf.org/mailman/listinfo/sidrops
>

-- 
Douglas Fernando Fischer
Engº de Controle e Automação