Please post the link to your TestMy results page. If you haven't already, read https://community.hughesnet.com/t5/Tech-Support/Think-you-have-slow-speeds/m-p/110034#M74607, as it explains what you need to provide in order to have speed issues addressed here.
Upload speeds of 2 to 3 Mbps are normal for satellite internet.
If you're on Beam 68, it's having some issues right now.
Apologies fr the delay but I am indeed on beam 68 and so can only rarely access web pages.
But >my< service is not my question.
I am asking how - in technical terms - upload speeds and download speeds can be so very different. Same beam, same satellite, same ground station(s), yes?
Yes. Speeds vary a lot.
When you connect to the Internet, the upload and download speeds you experience will vary based on a variety of factors including the configuration of your computer, the number of concurrent users, network or Internet congestion, the speed of the websites you are accessing, network management practices, and other factors.
Well, I'm thinking that within 1-2 minutes (such as when running a speed test using testmy), the configuration of my computer doesn't change, the number of concurrent users doesn't change, congestion doesn't change, website doesn't change....so the "variables" are close to unchanging.
What does change is the direction of the communication....
Actually, things do change, even though you think they don't.
"congestion doesn't change" -- oh, but it does, and constantly. Other variables that may not be on the list I provided change as well, possibly also constantly. Believe it or not.
Just to add to this, one way of thinking about it is that the beam is akin to a massive router with limited bandwidth that may be used at any time by any HughesNet user in Central Virginia; as well as any of the surrounding area that the beam covers. Similarly, the beam itself is combined with two other beams at the ground station in San Diego. Those beams cover parts of Missouri and Kentucky. On top of this, there's the terrestrial internet pipe to the ground station itself that could have bandwidth and/or routing issues (as has been known to happen). This example is specific to beam 68 and SDO, but can be applied to virtually anywhere in the network.
So there could feasibly be congestion anywhere in the system, concurrently affecting tens of thousands of users at any given time, depending upon the instantaneous user load, which is never constant.
With people on vacation (and 'staycation') and kids out of school this time of year that load is hardly predictable. Naturally this becomes more predictable as September rolls around.
And within that minute, 1000 subscribers may have hopped online, or decided to stream stuff, or whatever. Within that minute, HughesNet may have conducted a networking management action. Etc etc etc. Many things can happen in small periods of time.
It seems you're unwilling to believe these things, and unwilling to accept the explanations from those who have experience with this kind of stuff (Mark, for example). That's your choice, of course, but it seems a bit silly. The point is, the signal's behaviour is very unstable, even by the second. Accept it, don't accept it, whatever.
I understand that from hour to hour, day to day there are all these changes. But I'm asking about a time interval of no more than 1 minute....
It can change from second to second. At any time you can have people jumping online to download a file or stream something, and any number of them jumping on and off from one minute to the next. The more subscribers there are on a beam and/or gateway, the more subject it is to speed fluctuations, including minute to minute.
And at the end of the day the entire system is subject to this, as the two main satellites, including the one your'e on, have finite throughputs, though the individual beams are more susceptible due to having lower throughputs than the entire system itself.
It's one of the drawbacks to satellite internet. Unlike systems that are terrestrial based, there's no upgrading a satellite to increase its throughput, nor the individual transponders on it.