FYI - The speed issues with VPNs are:
1. The security protocols required can't be used in conjunction with the TCP/Web Acceleration techniques being used, so everything basically works at the uncompressed channel rate.
2. There's a lot of handshaking going on, so take your normal 600mS ping rate and basically multiply it for each packet.
As usual, the severity of how slow it can be is generally geometrically proportional to the level of congestion. If you want to get an idea of just how slow and irregular it can be, ssh into a server and try to type commands... it becomes predictive typing waiting for the echo after a while. lol
I think you misunderstood my comment about using VPN. I said using VPN worked better than without, not the other way around.
The reason for this is that the throttling HughesNet imposes based upon sites and content is circumvented when using a VPN.
You are again speculating without any hard data -- since HN does not reveal their optimisation strategies, all you are doing is guessing without any real data to support what amounts to basically an opinion. Not criticising you; you can speculate all you want, but it's nothing more than guesswork.
MrBuster wrote:
- That request they put out asking for user input about work/education related sites they are having trouble with would imply they are intending to give these sites extra priority -- throttling them would be very counter productive to the entire purpose of the effort. After all, they are not putting all this is place to roll out a bunch of new plans with crazy limits and glowing names, right?
Or they could be trying to find out which sites are more problematic, and why, in order to make various tweaks to the system for those sites to better work.
maratsade wrote:You are again speculating without any hard data -- since HN does not reveal their optimisation strategies, all you are doing is guessing without any real data to support what amounts to basically an opinion. Not criticising you; you can speculate all you want, but it's nothing more than guesswork.
Sure speculation and guess work -- I am sorry the wording was not clear. I do not intend to mislead anyone and I hoped the usage of words like 'appears' would express that. Some routers can show you detailed data usage per device.
So with router logs for every day's activities saved, I can see how much data a device pulled and when it pulled it, how fast it pulled it and the general latency in ms associated with the activity. If I may be so bold as to offer some speculation, it appears that VDS does limit the amount of data that a single device can pull from obvious recreational video sites, and that there is an increase in latency associated with the operation for some reason which may just be coincidence, but when I see that line go up when a device is pulling data and that the line for data amount looks suspiciously like it is leveled off and fairly constant, I just sort of assume "whew -- good thing that VDS is on!"
If I turn off the VDS, the device appears to pull data faster, and the latency numbers drop back down. The device also appears to consume more of my data as many streaming sites seem to jump to a higher resolution, but my eyes are pretty bad. Leaving the VDS off when the crew comes through, even for a brief time, and I usually see several GB disappear to some silly android phone. Of course I do not mean to imply I am sure of this.
It appears the VDS does what the name implies, as I think GabeU had some Fast.com results that showed results comparing data rates with the VDS on/off, but maybe he can clarify. Just by glancing at a graph I feel like I have a pretty good idea if my niece or her family is downloading a video from some site (maybe Netflix?), or streaming from someplace, and I just kind-of/sort-of guess at if the VDS is on or not without actually looking. I do not mean to imply that this is anything other than speculation.
Regarding the interpretation of the policy they posted, sure I could be very mistaken -- probably wrong. I just can't imagine why they would ask users to identify the sites they need improved if they did not want to do something to improve them. But then I am not very imaginative.
I think we did understand; it's just that VPNs don't work well for satellite internet, even though they may work for some people (as you demonstrate). For other people, VPNs slow down the network and don't work well.
Not sure what you mean with "the throttling HughesNet imposes based upon sites and content." HN does not throttle sites and content; they throttle speeds when users have used up their data allowance. EDIT: they also conduct network optimisation to improve user experience (though this is subject to congestion and internet latency), and while network optimisation practices are not revealed by the company, throttling is not one of them (except of data when the user has exceeded allowance).
grizzle wrote:I think you misunderstood my comment about using VPN. I said using VPN worked better than without, not the other way around.
The reason for this is that the throttling HughesNet imposes based upon sites and content is circumvented when using a VPN.
Ideally, it would be best to have various users on different beams to try out Teams with VPN and Teams without VPN. I am pretty sure I have been using Teams and Webex both ways, and it seems good both ways but trying under conditions of some congestion would be good to see if this is a factor. I have been pretty busy today, and with ping times flat lined near minimum for satellite until after 9am and staying insanely low until after 3:30pm, I don't think I have had a fair chance of seeing any congestion yet for a fair test. It operated great today, but today was unusually fast judging by the numbers I am looking at.
MarkJFine does have a good point. I have noticed that uploading to an FTP with the VPN can be painful, and fast if I kill the VPN. The reasons he mentioned are spot on.
Anyway, the "throttling" of video is through the VDS, and I am pretty confident of the following speculation regarding the HughesNet optimization strategies*:
*edit
