Subscriber uplink is likely using some sort of TDMA - those protocols are very difficult to keep efficient when you have hundreds of users in a channel, weather considerations, priority classes and 240ms latency from all clients to the headend. Adding 20-200ms for that is probably reasonable. Also you probably can't internally route on the network to go to the backbone link closest to the desired internet host - likely everything comes out of one satellite uplink location that adds its own latency to hosts that are located on the other side of the country.
On Fri, Aug 3, 2018 at 10:28 AM, <[email protected]> wrote: > Lesseeee 22,400 x 2 /186000= 240 mS one way. > So 480 mS ping RT under perfect conditions. > > But in real life it is twice that. No idea why. Perhaps they are using > another satellite connection for backbone feed. > > *From:* castarritt > *Sent:* Friday, August 3, 2018 9:22 AM > *To:* [email protected] > *Subject:* Re: [AFMUG] Satellite internet stats > > It is the ~23,000mi orbit geostationary birds that kill latency. When a > ping to a terrestrial server has to travel >90,000mi, the latency will > never be better than half a second. Musk wants to do a large constellation > of low orbit birds. > > > On Fri, Aug 3, 2018 at 10:18 AM Seth Mattinen <[email protected]> wrote: > >> On 8/3/18 8:03 AM, Bill Prince wrote: >> > Latency still sucks. 700-800 ms if there is no congestion. Multi-second >> > latency if there is congestion. >> > >> >> >> Time for some quantum entanglement. >> >> -- >> AF mailing list >> [email protected] >> http://af.afmug.com/mailman/listinfo/af_af.afmug.com >> > ------------------------------ > -- > AF mailing list > [email protected] > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > > > -- > AF mailing list > [email protected] > http://af.afmug.com/mailman/listinfo/af_af.afmug.com > >
-- AF mailing list [email protected] http://af.afmug.com/mailman/listinfo/af_af.afmug.com
