> On 5 Apr 2020, at 16:22, Dave Taht wrote: > >> > > I'd still be willing to bet, then, that the majority of instances were > not turning nat mode on, when > they should have been. If memory serves, at the time there was a lot of concern about cpu usage and (I felt) that every line, every potential instruction and cache hit was being scrutinised. NAT lookup by default was deemed too much. Cake ended up being ‘cpu heavy’ though I’ve yet to see a better combination of shaper, aqm, flow fairness, host fairness, DSCP awareness and ack filtering in one overall package, let alone one that did it in less cpu. Cheers, Kevin D-B gpg: 012C ACB2 28C6 C53E 9775 9123 B3A2 389B 9DE2 334A