[Bloat] [Cake] bbr vs all the aqms, cake winning...

[Livingood, Jason]

On 9/28/24, 21:35, "Bloat on behalf of David Lang via Bloat" <bloat-bounces at lists.bufferbloat.net <mailto:bloat-bounces at lists.bufferbloat.net> on behalf of bloat at lists.bufferbloat.net <mailto:bloat at lists.bufferbloat.net>> wrote:
> That's why I was saying 'traffic simulator' rather than 'traffic replay'. A good 
simulator is a much harder problem

[JL] Further to David R's point - there is so much variation in client, device, home network, CPE, etc. that simulation just never catches all the corner cases (and so IMO not worth the time investment in most cases). That's why in my experience large operators of networks and applications will of course run automatic QA tests in a lab, but quickly move on to phased deployment - with a gradually expanded deployment that is coupled with alerting for changes in operations patterns/performance/etc.