On 23/8/21 4:32 am, Gavin Troy wrote: > Using the same methodology (although with parallel=7): > > 16 17 1.06 > 8 8 1.00 > 78 78 1.00 > 6 5 0.83 > 23 23 1.00 > 4 4 1.00 > 17 16 0.94 > 30 30 1.00 > 8 8 1.00 > 8 8 1.00 > 1.00 avg > > So with my connection there's no significant gain or loss. Since this data > was uninteresting I tried again with a large packet delay. > # tc qdisc add dev wlan0 root netem delay 500ms > > 32 27 0.84 > 194 184 0.95 > 37 39 1.05 > 184 178 0.97 > 42 37 0.88 > 257 220 0.86 > 92 62 0.67 > 65 44 0.68 > 77 60 0.78 > 54 51 0.94 > 0.87 avg
Thanks. My takeaway here is there is no drawback to adding this patch, only a potential (average) gain for some people. So I am accepting this patch. Allan
