Forum Discussion
Speed tests
Even though I've never had such an issue myself, I've often followed the threads in tech support about speed problems. Usually in one of first couple of replies will be instructions about running a series of speed tests, and how they are absolutely necessary to trouble shoot your system.
Why are they necessary? Earlier in the week my speed was 30.24, this morning it's 27.90. What does this tell one, other than the system had less load on it on the first test?
Here's some speed tests that I ran within minutes of each other this morning:
Hughes speed test: 27.90
25mb file size with testmy.net: 37.4
Testmy.net default file size: 30.6
Perhaps that's why Hughes prefers the testmy.net 25mb test over anything else. It always reports much higher speeds on my system.
I recently followed one members travails with speed issues, who had testmy.net results going back over a year, which was not good enough. It never is! The tests have to be performed in an exact certain way. One file size for gen4, a different size for gen5, 3 tests, 3-10 minutes apart, but not an hour apart, 3 times a day, morning, noon, and night.
BTW the member's issues were not resolved as of today, after a month of running more speed tests. Of course, the member was told that the testmy.net 25mb tests showed that they were on the threshold of being within the promised speeds. Not surprising, if their system is like mine, where the testmy.net 25mb results are always about 10 mbps higher than the Hughes speed test.
Sorry, but all of this is beginning to strain my credulity about the usefulness of all these speed tests. Especially since I believe that Hughes may be cherry picking the tests for the highest speeds. Why else would they refuse to accept the results of their own speed test?
"I know I'm paranoid, but am I paranoid enough?"
Tom Clancy
It's not just the number that matters, it's the consistency. For example, If it starts out gang busters, but falls off and stays there for the duration, it may mean an MTU problem at the gateway. If it starts off then dies and starts up it might mean a problem upstream. It can also mean different things when seen during peak and non-peak times during the day.
These cases are really where the graph in TestMy is extremely helpful over other sources.
MarkJFine wrote:It's not just the number that matters, it's the consistency. For example, If it starts out gang busters, but falls off and stays there for the duration, it may mean an MTU problem at the gateway. If it starts off then dies and starts up it might mean a problem upstream. It can also mean different things when seen during peak and non-peak times during the day.
These cases are really where the graph in TestMy is extremely helpful over other sources.
Interesting. I have to admit I had to go to Wiki to find out what a MTU was, the last time I did network communications programing, was sometime in the last century before ethernet was common. I was using RS232 and DH+.
If it's a problem at the gateway or upstream, wouldn't all the users of that gateway be effected?
Can they switch you to another gateway? Do they have enough overlapping beams that they could even switch you to another beam?
I wondered if the actual shape of the graph had something to do with it I looked at a few of them from people that were having speed problems, and nothing jumped out. Maybe if one could overlay this graph with a graph of the loads on the gateway and beam?
It still bothers me that they're using a test that consistently reports 10 mbps more than their own speed test. Well, I've been bothered before. :>)>
gaines_wright wrote:If it's a problem at the gateway or upstream, wouldn't all the users of that gateway be effected?
Can they switch you to another gateway? Do they have enough overlapping beams that they could even switch you to another beam?
Depends on the path. 'Upstream' could be any hop along the way, not just the gateway provider. However, it is usually the gateway provider that determines the path because it picks the next hop in the chain.
Each gateway is assigned between 5-8 spot beams which are geo-fixed. If you're in the overlap of two spots they might be able to switch you to the overlapping one, which might get you a different gateway. It's also possible in some cases where they can get you on J1, which would also be a different spot/gateway. The latter would obviously also require realigning the dish.
gaines_wright wrote:Even though I've never had such an issue myself, I've often followed the threads in tech support about speed problems. Usually in one of first couple of replies will be instructions about running a series of speed tests, and how they are absolutely necessary to trouble shoot your system.
Why are they necessary? Earlier in the week my speed was 30.24, this morning it's 27.90. What does this tell one, other than the system had less load on it on the first test?
The speed tests help to determine not only where the problem lies, but also its possible cause(s). A set file size is used for the reason Mark gave, which is consistency, which Hughes' own test doesn't have. They also need to be run by the customer because not all speed issues are as a result of HughesNet, and only a customer running speed tests at their location can help to determine that.
gaines_wright wrote:Why are they necessary? Earlier in the week my speed was 30.24, this morning it's 27.90. What does this tell one, other than the system had less load on it on the first test?
Here's some speed tests that I ran within minutes of each other this morning:
Hughes speed test: 27.90
25mb file size with testmy.net: 37.4
Testmy.net default file size: 30.6Perhaps that's why Hughes prefers the testmy.net 25mb test over anything else. It always reports much higher speeds on my system.
Sorry, but all of this is beginning to strain my credulity about the usefulness of all these speed tests. Especially since I believe that Hughes may be cherry picking the tests for the highest speeds. Why else would they refuse to accept the results of their own speed test?
HughesNet's test is reliable enough for Tier 1 reps, but not reliable enough when troubleshooting the underlying cause.
gaines_wright wrote:I recently followed one members travails with speed issues, who had testmy.net results going back over a year, which was not good enough. It never is! The tests have to be performed in an exact certain way. One file size for gen4, a different size for gen5, 3 tests, 3-10 minutes apart, but not an hour apart, 3 times a day, morning, noon, and night.
BTW the member's issues were not resolved as of today, after a month of running more speed tests. Of course, the member was told that the testmy.net 25mb tests showed that they were on the threshold of being within the promised speeds. Not surprising, if their system is like mine, where the testmy.net 25mb results are always about 10 mbps higher than the Hughes speed test.
I'm reasonably sure of who it is you're referring to, and when the person can get some tests in using the protocol, which is designed and required for specific reasons, some of which have already been mentioned in this thread, the person will receive help. The tests already run are either too far apart, too close together, and/or there are too few during a testing "session".
"I'm reasonably sure of who it is you're referring to, and when the person can get some tests in using the protocol, which is designed and required for specific reasons, some of which have already been mentioned in this thread, the person will receive help. The tests already run are either too far apart, too close together, and/or there are too few during a testing "session"."
Well, in her latest post she said after some hardware problems were fixed, she had two days of normal service, but it didn't last. I personally think this was probably caused by the "infant mortality" of one of the components that was replaced. Does she need to run a new set of speed tests to find out? :>)>
gaines_wright wrote:"I'm reasonably sure of who it is you're referring to, and when the person can get some tests in using the protocol, which is designed and required for specific reasons, some of which have already been mentioned in this thread, the person will receive help. The tests already run are either too far apart, too close together, and/or there are too few during a testing "session"."
Well, in her latest post she said after some hardware problems were fixed, she had two days of normal service, but it didn't last. I personally think this was probably caused by the "infant mortality" of one of the components that was replaced. Does she need to run a new set of speed tests to find out? :>)>
My mistake. I thought you were referring to another member. One who had Gen4, then upgraded to Gen5, and had more than a year's worth of speed tests.
I was under the impression that the person you are now referring to was a new customer, not only from memory, but due to the 20 days of data resets mentioned, which only new customers, or those upgrading to Gen5 from legacy systems, get.
Thanks to all that replied!
I'm still dubious about these speed tests. I recently did a topic search for the word "slow" and then "speed" in the messages here. I then went to the solution, where one existed. I can't say I'm much wiser by doing so. I only did about 25 of them before giving up.
One guy obviously had a hardware problem since he had a 40% packet loss when a connectivity test was run. Another guy was throttled, which a quick look at the system control center would have told him. Another seemed to have his problem solved remotely, but I couldn't figure out what was actually done.
I know the instructions for doing all these speed tests are engraved on the stone tablets containing the Hughes help desk script, but can anybody point me to a thread where someones speed problems were actually solved by doing all of these finicky to perform speed tests?
"Does history record any case in which the majority was right?"
Lazarus LongThe tests don't solve it absolutely, but as I mentioned above, they do provide hints as to what the problem could be - aside from the obvious evening congestion issues (which I still maintain are from a majority people trying to stream TV at the same time). It's process of elimination.
