This is not a latency issue and is not consistent with HN speed tests.
My system: Win7, i7, Firefox, Bitdefender
"Modem" Signal: 111 Web Accel is on.
HN speed test: Down: 137, Up: 0.42
Web page load times using HN has always been slow for me. Now it is ridiculous! For example, the time from click to the time MyHughsnet.com loaded was *TWO MINUTES*.
A power off reset did not help. In the bottom left browser information display, most time is spent "waiting for XXXX", and some "transfering".
In no way does this reflect the transmission speeds indicated by HN.
What can be done so that my HN connection speed will exceed dial-up?
Woody - KZ4AK
I certainly wish HN would fix the problems! Service tonight has gone from awful to hardly able.
The waiting times remain the same. It can take quite a while to just do a TLS handshake exchange.
To add insult to injury, the download times dropped to 1 meg tonight per testmy.net.
Even timed out once loading email...
Raining, but signal strength hovering around 75 - 80.
Here is my sat data:
Woody - KZ4AK
Hello Felicia, et all...
Thanks for commenting... Last night (even very late) my connection was so poor that it was essentially unusable. I received a text message indicating a service tech is scheduled for 26 DEC.
However, I don't think there is anything service can do on my end. This morning, my speed is fine and the extended "wait" times (no data exchange activity) are minimal.
testmy.net results - https://testmy.net/db/KsjFpWAGN
Since nothing has changed at my location, it would seem the problem is not here, but resides with HN.
Given the circumstances, how do you think a service call would help?
Thanks for asking. I wish I could say yes, but web load performance is still poor. Even with ethernet to modem, a 110 signal strength, and great speeds reported by testmy.net, it sometimes takes a minute or more to load a very simple (minimal images) web page. Any improvement is too subtle to notice. I will say that the issue can vary. Sometimes it isn't really what I would expect, but is OK, at other times, seems like half my usage is spent "waiting". CNN.com just loaded in 15 seconds - Right now it is good, not so much earlier tonight.
Poor overall performance is still poor. Signal >100 testmy.net = 40 MB and faster.
Experiencing long download times that should download very quickly.
For example, just updated Mozilla Thunderbird. Watching the download progression readout, was running a paltry zero point one (0.1) meg bytes /second. Downloading a simple 1.4 meg data file can also be equally sluggish.
Why is the "actual" download stream so much slower than what should be, given the reported speed?
If I could attach one, a video would show the issue, albiet a long, boring video 😉
There was another tester on HughesNet, whose name I forget, who was getting a lot of wacky speed test results, as well. Pretty much any test showing you a result higher than about 50Mbps or so is "wacky", or incorrect.
If you look on your Results page and click on the TID of some of the wacky tests, you can see in the graph that something isn't right. When they start off at nearly 300Mbps, then ease down to something lower, you know something's a bit off. I don't know if it's the server or what, but something's definitely weird.
Edit: I should add that, for the really crazy ones, the graph often isn't there. Maybe the tests are too messed up to produce one.
Interesting, but it does not appear data speed / testing is the issue. I ran several ipconfig tracert passes. I attached two results. I understand the 600-700 ms latency, but not the timeouts. Lots of hops too...
The provider for your gateway is Level3 (22.214.171.124).
Many times you'll hit a hop that's isolated from being pinged in a traceroute and you'll see a timeout from that. However, in this case, I'm pretty sure Level3 are timing out at some point between when each reporting ping goes out and comes back, which is why it doesn't seem consistent. I want to think it's an MTU-related problem for them, but that almost seems too simple a fix. Likely also messed up routing tables that end up going to dead hops - they're known for that too.
Not gonna lie, Level3 is a mess and the source of a lot of the problems most of us see, not HughesNet. This is not only with regular browsing, but the source of streaming problems as well.
Edit: FYI: That second image shows shows you going into the Amazon Web Services 'black hole', which QRZ are hosted on.
Thanks Mark, but what can be done? Never had this issue prior to joining HN, when was with Verizon 3G. When the local 3G tower transmit signal strength became intermittant and Verizon could not (would not? - they were loosing interest in 3G) find out why, I changed my ISP. Yesterday, could not even download my email from Dot5 Hosting. Coincidence?
What can be done?
Oh, and I agree... AWS seems to be the slowest of the slow...
Also apparent with their video streaming service. Maybe they do everything on a single server.
My issues with AWS are really security-related. Seems they cater to quite an adventurous bunch on their hosting side, causing me to become Blocky Balboa: 34 .htaccess "Deny from" lines of single and netblocked IP ranges from AWS. Almost as much as I have for China.
As for what to do... we're on the same beam (68 - reading back) so pehaps @Amanda or @Liz can raise the issue to someone in networking, so they can let Level3 know to fix their erratic San Diego gateway connection. Whether that's successful or not, we're still at the mercy of Level3's level of competence. #fwiw
If the traceroute was timing out in spots, the pings would definitely timeout. Traceroutes are basically incremental pings to each hop along a route. One ping to the end destination is going to inevitably timeout at the hop that was timing out during the traceroute.
Did some cursory checking here during a relatively light period, and it looks like the Level3 server handling SDO068 (and possibly all of SDO) is fubar.
Everything, and I mean everything is timing out right after that hop.
I can believe that! Question is, how to fix the issue. I contacted phone support and *finally* got them to involve "engineering". Eng was to call back, but after over two days, nothing from them.
Not to worry. I think things are happening.
But like I said, it's a Level3 issue. All HN can do is poke them in the eye.