Hughesnet Community

The all to familiar problem, slow - slow - slow

cancel
Showing results for 
Search instead for 
Did you mean: 
hamradio
Senior

The all to familiar problem, slow - slow - slow

I have complained about the throughput speed before and there was no real solution  Sometimes downloads are near what one would / should expect, but most often very unsatisfactory.    No way for me to prove it, but all evidence available points to Hughsnet gateway bottlenecks.    Sometimes the download speed may be (sort of) OK, but the actual throughput is usually very slow due to "waiting"  - periods of inactivity in responding to a request.

I have a 101 signal strength, plenty of data, recently power-off reset the modem, Ethernet connection to moder, ect.  Near dialup true throughput persists in spite of the indicated fast speeds by testmy.net.  

Also, Testmy.net (recommended by HN) does not seem to be reporting acturately.   The most recent run indicated the speed topped out at 98Mbs. It has reported speeds in excess of  200 Mbps Yeah, right! 

https://testmy.net/db/62oQU7mpa      Here are others:  https://testmy.net/db/UDZZeu0Q3   https://testmy.net/db/-aL2Zg~-R

A test by speedof.me (HTML5 Test) produced results that reflect what throughput speed I actually receive:

https://i.speedof.me/190925232454-2087

Sometimes the crazy high latency reported in this test are more a more realistic value, around 660.  However, the speed is still slow.

My question is will HN ever fix the problem that I (and many, many others) are plagued with?

 

Thanks,

Woody  -  KZ4AK

 

 

 

https://i.speedof.me/190925232454-2087

85 REPLIES 85

Yet another thing...  testmy is very often "losing connection", then "reconnecting".  Sometimes multiple reconnects attempts... Quite a bit of that during one session a few minutes ago...

Here is another scn grab of results...

If I use the link (  https://testmy.net/db/Bm5VlNtRx  ), I get a report, but it is quite different from the first screen data when the test finishes:

Clipboard04.jpg

GabeU
Distinguished Professor IV

@hamradio 

 

What that picture shows makes absolutely no sense.  All three things should coincide, as in the result showing at the top, the graph from the test in the middle, and the result showing as the most recent one below.  That the graph is WAY off is crazy.   

 

I wonder if the crazy graph on the test page is reflecting the connection issues, while the graph of the finished result, as in when you click on the TID for the test on your Results page, is eliminating those connection issues and is only showing the results from when the test was legitimately running?  

 

This...

Capture1.JPG

 

or this...

Capture.JPG

 

Either way, it's an indication that something is VERY wrong.  

 

I'm sure it's good that you're demonstrating all this, as the reps/engineers should know it.  

I think I remember that we're going to the same gateway, and using the same nasty connection via CenturyLink.

I think what you're seeing are the dead gaps that CenturyLink is inserting, and it's showing up as very strange catch-and-release spikes in the tests.


* Disclaimer: I am a HughesNet customer and not a HughesNet employee. All of my comments are my own and do not necessarily represent HughesNet in any way.

Thanks guys,,,

I certainly agree that something is weird.  I hope  HN engineering will/can figure it out and correct it.   FWIW, performance isn't too bad right now (11 AM Eastern, 10-17-19.   As mentioned, sometimes the system works, just not too often or for too long.

_W_

 

FYI, One more pair of scn grabs.

First is the scn returned after the test.  Second is what I get after clicking the TID.

_W_

Clipboard01.jpgClipboard02.jpg

@hamradio,

 

Thank you for providing me with these tests. I will now be escalating a case for you. Please allow 1-3 business days for our advanced technical department to reach out to you. I will also be private messaging you the case number to use if you call in before they reach out. 

 

-Damian 

Got it - Thanks Damian !    Will be waiting to hear.

 

Observation:  Earlier, I was asked if I could see a reference to HN on testmy.   Yes and no...   No mention of HN when I login to testmy.  However, after a test, if I scroll to the bottom of the page, there is a clickable box for HN.  Clicking the link sends me to a page that mentions HN and has links for  up / download tests.  I tried it that way and noticed no difference ( https://testmy.net/db/VktXVI1c6  )

-  The initial test results page is still quite different from what is seen by going to the link -- still reporting the wacky, beyond reality, Mbps maximum. 

Testmy is also still loosing connection and reconnecting - repeatedly.  

_W_

I received a call from HN tech support.  They reported a number of issues similar to mine from my area.  A visit by a field tech has been scheduled for 22 October.   They indicated the tech may change my dish polarization.   I understand antenna polarization (e.g. vertical/horizontal/left-right circular), but uncertain what the implications may be for my connection.  The signal strength is typically =>100...

Unless this puts me on a different beam, not so hure how changing the polarization will fix the data delay issue.

I guess we will see....

_W_

Rain in the forecast for the 22nd, so rescheduled for the 23rd.  

Will report back with visit details...

Woody - KZ4AK

Although I had rescheduled for tomorrow, the HN contractor came today (22nd).  Per the work order, he changed the antenna polarization.  Given my symptoms, he was no more optimistic than me about the "fix" working.   ...But certainly giving it a try!

Will need to gather some data over time to see if performance actually improved.  

In a day or two will report back with observations.

Woody - KZ4AK

 

Results of polarization change as of 23 Oct....   Viewing small youtube videos, on a week day morning seems to incure less buffering.  Not so much for evening times - still significant buffering.  The % buffering time is a bit hard to quantify.  Still evaluating.

Here is an example of performance on a week day evening...  Time to download (not install) a Microsoft update  of 6 files totaling 385 MB was 15 minutes.  FWIW, for the first 3 minutes after starting, there was zero download activity. 

If I got the math right:

385 MB x 8 = 3080 Mbits. in 900 seconds,  3080 / 900 = 3.42 Mbps.  If I got this right, this is still nowhere near the speeds reported reported by testmy.

After more evaluation will report further....

Woody - KZ4AK

No improvement yet.  Placed another call to phone tech support.  They called back Sat, but after hearing my symptoms, wanted to wait until the begining of week when higher level support will be available, then they said would call me.   So, waiting to see.

_W_

GabeU
Distinguished Professor IV

@hamradio 

 

One thing you've got to remember is that system updates, especially from Microsoft, are one of the worst ways one can gauge the speed of their service.  Microsoft downloads, whether for system updates or even just program downloads from their servers, can be awful, regardless of your overall speed, and even if your speed was great and there was virtually no congestion to speak of, though the ones through Windows Update tend to be the worst.

 

I had a ~600MB update a couple of years ago that took over 26 hours.  It's the only time in my nearly 15 years with HughesNet that I've left a computer on overnight.  Updates were always a bit slow around that that time, but never THAT bad.  They time/speed eventually crept back up to what had been the norm, but even to this day they sometimes take longer than you'd think they should.  

 

Downloading the updates directly using WSUS or the Update Catalog can be faster, but they can still take a long time with those methods, as well.  

 

To be honest, though downloads of any kind don't always tend to match overall speed test results, when I need to verify real time speeds I've found that downloads from Linux Mint tend to be the best for me.  Usually from Clarkson U or James Madison U.  I don't let if finish, of course, as a "test" would use upwards of 2GB each time, which would be ridiculous, but I let it run for at least twenty or thirty seconds to see what's going on, then stop it.   It can use more data than a testmy test, but it's still a decent way to gauge overall download speed in real time.  Again, though, just like other downloads, it's not always the most accurate for that purpose, as you're relying on servers that aren't designed for speed tests, but rather downloads, but it can still give a good idea of how your service is doing in a real world scenario, overall speed wise.

 

Regardless of this particular aspect, I hope they can remedy your issue(s) permanently, and soon.


@hamradio wrote:

No improvement yet.  Placed another call to phone tech support.  They called back Sat, but after hearing my symptoms, wanted to wait until the begining of week when higher level support will be available, then they said would call me.   So, waiting to see.

_W_

 

No call back on case 124537656.  So, I called them.  This is SO frustrating!!!!   The 1st level support person said an attempt was made to call me.  NOT true.  No calls, no missed calls in phone log.  Then told me a speed check indicated no problems and the case was closed.  I told her that a simple speed test would not show the problem and that is why I needed to escalate to engineering.  She did not understand, repeatimg the same old mantra about speed tests and latency.    Also, said that since speed tested OK, she could not revive the old case or create a new one.  Asked for supervisor.  At first the supervisor focused on one pdf file download I used as an example.  It took a lot of conversatiom, but finally convinced her that the pdf download was not the central issue, but only an example.of a general problem. I repeatedly told her that a simple speed test would show all ok, but would not detect what is happening.  She spoke with engineering and returned with the usual -  Since speed tests were within specs, there was no promlem and so, would no open a new case.  After multiple repeats, and at my insistance she re-contacted engineering.  This time she said engineering would call me in an hour or so.   I hope they really do.  It seems that HN does not have the ability to comprehend the difference between instaneous speed checks and overall network throughput - Or they will not admit to having a problem, or both.  The supervisor asked me to do another testmy speed test.  She was at a loss as to why the download speed result was 266 Mbps. 

https://testmy.net/db/8sU3mOZG7

Woody - KZ4AK

@Liz

@Dameon


 

Gabe,

"One thing you've got to remember is that system updates, especially from Microsoft, are one of the worst ways one can gauge the speed of their service. Microsoft downloads,... " 

 

Ok on that.  It was just another attempt to provide an example.  Like the multiple failures to download a pdf - the example I used with phone support. ...After several tries it finally downloaded. 

 

Due to phone support's lack of understanding of the difference between and example and the general problem,  phone support initially focused on me having only one problem, downloading a specific pdf.   So frustrating trying to explain that the pdf was not the base issue, just an example.

BTW, I know you understand the difference but were simplty commenting on the specific example.

_W_

 

FWIW, here is a graphic of what my symptoms indicate  (and what HN does not admit to, or test, when trouble shooting a throughput issue):Untitled-1.jpg

Damian - Sorry, got you name incorrect in an earlier post.

Have gotten only one thing from engineering, disappointment.   The proper issue has not been addressed even though the "case" was closed, and they are failing to call back as promised.   They simply did a remote speed and system check, then concluded there was no problem.  They did not  seem to understand the problem is with poor throughput due to delays, not speed. It seems (although no one will address it) the bottle neck is the result of problems within HN;  hubs - network op center - outside world.     Sure appears HN has over sold their capability to deliver, or there is an undiagnosed internal problem.  See my recent posts.

Woody - KZ4AK

 

@Damian

@Liz

The top part of that red box is CenturyLink and I'm telling you it's them, and they're getting worse.

I had all three twitterbots timeout again yesterday and none of them had anything to do with HughesNet, just the CenturyLink backbone. It's broken.


* Disclaimer: I am a HughesNet customer and not a HughesNet employee. All of my comments are my own and do not necessarily represent HughesNet in any way.
GabeU
Distinguished Professor IV


@MarkJFine wrote:

The top part of that red box is CenturyLink and I'm telling you it's them, and they're getting worse.

I had all three twitterbots timeout again yesterday and none of them had anything to do with HughesNet, just the CenturyLink backbone. It's broken.


It's unfortunate that a single company can create such a headache for so many.  Many smaller businesses as well, I'm sure.  They REALLY need to get their act together.  

I agree, but my only point of contact is with HN.  They need to apply pressure to CL to get the problem fixed, I cannot.  Engineering did not call back today as promised.   Apparently they don't care, so I have contacted the "Fed Com Comm" to see if that will get their attention.  Don't know what else to do. 

--Interesting the the abbrevation for that agency is not allowed on the forum - would not post if it was in content. Had to use "alais"--

 

Reply to me from Fed Com Comm :

--

Hi Norvell,

Thank you for your submission. Based on our review, your Ticket No. 3621612 was served on your provider on Oct 28.

Here's what you can expect from the "commission"s informal complaint process:

- Your provider should contact you directly in an effort to resolve your issue.
- The "commission's" role in this process is to facilitate a conversation between you and your provider.
- Your provider is required to submit to the "commission" a written response regarding your issue no later than 30 days from today.
- The "commission" will not contact you until we receive a response from your provider.