Jump to content


 


Register a free account to unlock additional features at BleepingComputer.com
Welcome to BleepingComputer, a free community where people like yourself come together to discuss and learn how to use their computers. Using the site is easy and fun. As a guest, you can browse and view the various discussions in the forums, but can not create a new topic or reply to an existing one unless you are logged in. Other benefits of registering an account are subscribing to topics and forums, creating a blog, and having no ads shown anywhere on the site.


Click here to Register a free account now! or read our Welcome Guide to learn how to use this site.

Photo

Speedtest vs ISP speedtest


  • Please log in to reply
10 replies to this topic

#1 j1983

j1983

  • Members
  • 57 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:HK852
  • Local time:04:52 PM

Posted 05 July 2016 - 05:34 AM

I've recently switched to a 1000M plan with a local ISP, I know results may vary a little bit and I don't expect to ge the full speed, but my recent speedtest results have shown 300-600Mbps depending on the time of day that I run the tests, which is not even close to what is advertised.

Actually, in the four and a half days I've had the service I've only tested once where it went over 600M, the rest of the time it's around 290 to a little under 400M when connected directly to my laptop.

 

The ISP has their own testing site which downloads a small 600kb file and runs it on your computer on that black screen window, while opening up a new tab in the browser to show results. There are separate tests for their 100m service, 500 &1000M as well as a separate one for Mac users. 

I don't know what type of test this is but it shows the speeds are 850-950Mbps. That's between 200-300+Mbps difference with their site and speedtest.net 

 

I've since tried Fast.com, Speedof.me and a local internet exchange site, all showing similar speeds which are several hundred Mbps lower than the ISP site.  Today the independent site tests were between 290-420Mbps while their site tests remained constant around 820-860Mbps.

 

I've talked to their customer service who were here today on a separate issue but they were no help and instead said they would have their I.T. guys contact me.  I don't know much about this stuff, I almost got a nosebleed trying to read multiple articles to find out about this earlier today but I'd like to try and understand this better before I speak to the company again. 

 

Thanks in advance


Edited by j1983, 05 July 2016 - 05:37 AM.


BC AdBot (Login to Remove)

 


#2 Trikein

Trikein

  • Members
  • 1,321 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Rhode Island, US
  • Local time:03:52 AM

Posted 05 July 2016 - 06:22 AM

Who is your ISP? Also what network equipment is being used?



#3 michelmau5

michelmau5

  • Members
  • 24 posts
  • OFFLINE
  •  
  • Gender:Male
  • Local time:09:52 AM

Posted 05 July 2016 - 07:24 AM

Youre maximum speed really depends on where you live even when your ISP says they can give you 1000m.

If you live very close to the exchange (I don't know if thats the right word i'm dutch) your internetspeed will be high (like 500m) but lets says you live like 3km from the nearest exchange your speed will be much less (like 100m).

Never trust your ISP's speed test.

 

Speedtest.net is the best option imo



#4 j1983

j1983
  • Topic Starter

  • Members
  • 57 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:HK852
  • Local time:04:52 PM

Posted 05 July 2016 - 07:32 AM

Thanks for the quick response.  It's called HKBN, one of three main ones in HK where I live.  I don't know about network equipment but I'm on a 1000M fibre broadband line.

 

 

Here's their site for the speedtest:

http://www.speedtest.com.hk/

 

You don't need to sign in or anything to use it, in case you want to try it out :)



#5 Trikein

Trikein

  • Members
  • 1,321 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Rhode Island, US
  • Local time:03:52 AM

Posted 05 July 2016 - 10:08 AM

Not familiar with non US ISP's. However in general, a ISP just provides you the connection to the internet. They can't control the speed of the internet as a whole. So a speed test on their network tells you how fast THEIR network is. A test outside their network can be effected by routing/congestion problems between your ISP and where ever the test is.

 

As for speed tests, not a fan of speedtest.net. Between the heavy advertisements, congestion from being a popular site, and depending on flash, I don't find it very accurate. 



#6 j1983

j1983
  • Topic Starter

  • Members
  • 57 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:HK852
  • Local time:04:52 PM

Posted 06 July 2016 - 02:49 AM

I don't expect speedtests to be 100% accurate but in the past I've always had them show results close to what I was paying for.

 

As someone who knows nothing about all this networking stuff, could I further ask what good it does me to have a 1000M connection to their servers but much lower reaching others?   One of the speedtests I did with the local HKIX was also the same one they claimed their test went through but still showed a significant difference in result.  

 

 

 

Maybe my issue here is an advertising claim but it seems they have the consumer by the balls with this technicality. 1000m though their network but 200-400M everywhere else.



#7 Trikein

Trikein

  • Members
  • 1,321 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Rhode Island, US
  • Local time:03:52 AM

Posted 06 July 2016 - 08:51 AM

The point of having a fast connection to your ISP is so your ISP connection is never the weakest link. It's like a car. Even if you can't go over 55 on most roads, if you find a road you can, you want to go as fast as possible. If you find certain tests are slower then others, find out what server you are testing to and run a tracert to it. There are "some" issues your ISP can help with if the problem is a routing or congestion issue in their network or on the hand off to the internet. Various tests can show where you "might" have a problem, but doesn't tell much on what or why. 



#8 j1983

j1983
  • Topic Starter

  • Members
  • 57 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:HK852
  • Local time:04:52 PM

Posted 19 July 2016 - 05:26 AM

So my basic understanding from this is that their test don't necessarily reflect real world speeds from the rest of the internet?

 

Forgive my continued ignorance on the matter but I'm trying to argue that I rely more on independent sites like speedtest, which I understand are used as a reference only, rather that believe the crap from their own website that shows a favorable result everytime.  I don't expect the full 1000M but even other factors like server locations and other stuff that may interfere with the signal can't cause a 70% difference in speed advertised vs the actual provided speed. 



#9 Trikein

Trikein

  • Members
  • 1,321 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Rhode Island, US
  • Local time:03:52 AM

Posted 19 July 2016 - 02:09 PM

"believe the crap from their own website that shows a favorable result everytime"

No, that is not what I am saying. What I am saying is the on ISP test tells you what you are getting on your ISP. It's two data sets. Think of it like the speedometer on your car. It says it goes up to 100MPH, but that doesn't mean you will always been going 100 miles in 1 hours. There are other things that slow you down when driving (traffic) so you can take other tests  that will tell you your speeds to different places. It's not a matter of accuracy, because you don't have "one" speed. It is dynamic. 



#10 j1983

j1983
  • Topic Starter

  • Members
  • 57 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:HK852
  • Local time:04:52 PM

Posted 19 July 2016 - 09:42 PM

Thanks again, the car analogy is starting to sink in, sort of. So the differing speeds mean the car is taking different routes, or going to different/separate destinations?  In any case, they promised they car would go 100Mph, it only goes 20, regardless of the traffic and obstacles that's a pretty bleep speed.

 

I was finally able to learn and do a tracert like you originally suggested, not that I understand what they mean. It took me ages to find out how to run one and it will probably take twice as long for me to figure out what it means which I'll try to attempt later when I get back home. I've posted two results in case you're interested in having a look. 

Tr, speedtest.com:

Tr2%20Speedtest%20from%20pc%20via%20wifi

 

Tr, to the ISP:

tr1%20hkbn_zpsn6wnqtwa.png
 

 

This is the ISP website, not the site they use to do their own speedtest which opens up an IPERF (whatever that is) test that opens up it's own window. 



#11 Trikein

Trikein

  • Members
  • 1,321 posts
  • OFFLINE
  •  
  • Gender:Male
  • Location:Rhode Island, US
  • Local time:03:52 AM

Posted 19 July 2016 - 10:47 PM

"In any case, they promised they car would go 100Mph"

 

Do they? A ISP just provides you your connection to the internet. They have nothing to do with how fast you go once you are on the public net. Think of them as providing your driveway, or the roads to the highway and the highway being everything else. Using your first tracert as a example, you're on your ISP network until hop 6. Notice it reaches it destination almost instantly after. That is because the test server is in a data storage center directly on the ISP. In your second example, notice the huge latency increase from hop 5 to 6. That is likely caused by congestion inside Telia, who is acting as transit between your ISP and Cogentco. Think of Telia being like the on-ramp onto the highway and Cogentco being a toll road your ISP pays to offload traffic onto. I think the problem here is network management problems by Telia, like seen here. 

 

:;Edit:: I think you should contact your ISP and show them that second tracert. I would suggest contacting their social media team or looking for their personal forum to see if it is a trend. They probably won't/cant do anything directly to fix the issue, but if enough people report it, it will be escalated to engineering just to avoid call volume. You will probably get push back. Just be friendly but persistent. 


Edited by Trikein, 20 July 2016 - 02:11 PM.





0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users