8 replies to this topic

[kenwood850]

Up until recently I never bothered to check my Internet speed on my ADSL connection but over the past couple of months I noticed that at certain times (early evening mostly) my connection to the Internet seemed sluggish.  I will not bore you with all the details but one evening when it was particularly sluggish I checked the speed with a well known test site and found my speed was only about 15% of what it should be.  Since then I have been working with the ISP to resolve this but in the process I found testmy.net and now have two questions.

1.   Even during the best of times when everything seems fine I only receive about 90% of the speed I am paying for even when using the ISP’s flash speed test site.  When I mention that speed value to tech support the say this is a normal and acceptable condition.   Is no more than 90% of the rated speed all you should really expect?

2.   When I do a speed test using the ISP’s flash speed test site and it reports 90% of the my connection speed, I often find that if I immediately do a speed test with testmy.net it reports a speed considerably less.  I have read the sticky postings and some of the other postings on this subject and understand why this is the case and how the testmy.net results are more real world and meaningful, but how do you convince your ISP of that?  All the low speed readings in the world from a dozen different external speed test sites that show a problem will not make your case if the ISP only wants to judge your connection speed based on their internal flash speed test site that shows everything is 90% of fine.   Has anyone had any success in convincing them otherwise and if so, how did you do it?

[mudmanc4]

Welcome to testmy.net forum

I will speak to the " convincing the ISP " question , depending on the ISP , the customer service will vary greatly. Many have very qualified knowledgable technicians available, yet as a general rule , when someone calls about a slow connection , or of the likes , you immediately get the as someone recently stated " rolodex run around ".  I like that statement because nothing could be more true. A series of standard questions run along side the same series of answers. Simple put , there are so many people that have issues , and the lot of them finally call the number , problem is , a large percentage haven't a clue what they should expect to hear from customer service.

The next issue is customer service expects the same general questions , as there plenty aware of the service there providing and the limitations involved. What I mean to say is the ISP knows what the quality of service they will give to there customers , as they themselves are contracting bandwidth , many times smaller ISP's will buy from larger providers. Now everyone would like to sell as many connections as possible , this creates more issues. Getting to the point.

Think in other terms when calling your ISP about speed issues. Try and think as if you are selling something to someone, and what the responses you would give , if you knew there really was no better answer to give. Never ask a question (directly) you can ask questions in many ways ) that you don't already know the answer to. Think several questions ahead of the last one you asked , put them into a corner by waiting for them to respond. Do not get upset , do a little homework before calling ( your reading this which means your on the right track already )

Pay close attention to the testmy.net database of ISP's and there results overall. The data stored at testmy.net is compiled over long periods of time and hundreds of ISP's around the world , which makes it a great place to do a little more digging into the past present performance of your ISP in general.

Call your ISP ( repeatedly ) and when they ask if you would like to have someone come out and test , always say yes , of course there are those that charge to have a tech come out ,  and in some instances you are left with no choice if you would like to get results. This does several thing , first and most important , it creates a record of your issues. Second , with that data already in there systems , you can use this further down the road to get to a higher level technician and actually get to the point of the issue.

Make sure to test regular , as a member of testmy.net your testing records are stored indefinitely, ( unless of course you decide to delete them ) you can use this information to show the technician your connection averages, if they claim they cannot guarantee speeds outside of there network , this is true in many ways , but in others , they have much more control then they might lead on. After all , most businesses are in it for the money , and the less time they spend on technical calls to a higher paid employee the better.

Just always be kind to customer service , it usually gets you further. Educate yourself by reading through others issues in the testmy.net forum, see what solved there issue , find out what there ISP said to them and what the outcome is / was. Asking educated questions and giving educated answers that an entry level customer service might not be well versed on , or is not at there disposal to answer , will likely get your call elevated to a 2nd or even 3rd level tech. Which in turn raises your chances of looking further into the issue and your quest , getting what you are paying for.

[CA3LE]

kenwood850, on 29 October 2011 - 06:30 PM, said:

but how do you convince your ISP of that?  All the low speed readings in the world from a dozen different external speed test sites that show a problem will not make your case if the ISP only wants to judge your connection speed based on their internal flash speed test site that shows everything is 90% of fine.   Has anyone had any success in convincing them otherwise and if so, how did you do it?

First, welcome to my website.

One thing you'll want to do is run a traceroute to the speed test server they give you.  It's most likely hosted within their network.  They say that this is to eliminate the variable of the other guys server or traceroute.. blah blah.. BS, they probably don't peer very well.  The basically have you test on a server within their network and eliminate the variable of... the internet.  I mean seriously, how can it be an internet speed test if you're not even extending the connection beyond the host.  I think it should be illegal for ISPs to do this to their customers.  I think they should be required to use a third party testing site.  But it's become common practice, I'm trying to change this.

You may want to run a test on the Washington D.C. server.  See if your results vary.  http://dc.testmy.net (this will change your default server to DC, links to change back are on the test pages... or clear your cookies)

I have very powerful servers and one of the most well connected hosts in the world.  Chances are, TDS is just not delivering and hiding the fact with the internal test.  You can feel that it's not right, otherwise you wouldn't be here...  

So what speeds are they supposed to be giving you?

Another warning, the server address you go to when you perform their speed test isn't necessarily where the test is coming from.  They could host the site it outside of their network.. but transfer using an internal server.  Tricky business and it should be against they law.  As I've said many times, you're paying for Internet not Intranet.  It's just not real world.

[RTB]

Actually, having both an internal and an external server lets you find out more specifically where a problem might exist. If the internal server is slow to begin with, the problem is local. If it's fast, but the external server is slow (might have to test at multiple locations) then there's a problem with peering. Of course... this is assuming such a bog-standard flash test bothers to do an accurate test. Of course, a fast internal test only says something about internal speed, and claims that it's an accurate representation of browsing speed are merely lies.

The reason why 90% is considered perfectly fine is that sending data over the internet has overhead. There are headers/footers added to data as it's being sent/received, and those count in the maximum throughput but not for download speed, because they're not part of the data being sent/received.

[TriRan]

mudmanc4, on 29 October 2011 - 08:04 PM, said:

Welcome to testmy.net forum

I will speak to the " convincing the ISP " question , depending on the ISP , the customer service will vary greatly. Many have very qualified knowledgable technicians available, yet as a general rule , when someone calls about a slow connection , or of the likes , you immediately get the as someone recently stated " rolodex run around ".  I like that statement because nothing could be more true. A series of standard questions run along side the same series of answers. Simple put , there are so many people that have issues , and the lot of them finally call the number , problem is , a large percentage haven't a clue what they should expect to hear from customer service.

The next issue is customer service expects the same general questions , as there plenty aware of the service there providing and the limitations involved. What I mean to say is the ISP knows what the quality of service they will give to there customers , as they themselves are contracting bandwidth , many times smaller ISP's will buy from larger providers. Now everyone would like to sell as many connections as possible , this creates more issues. Getting to the point.

Think in other terms when calling your ISP about speed issues. Try and think as if you are selling something to someone, and what the responses you would give , if you knew there really was no better answer to give. Never ask a question (directly) you can ask questions in many ways ) that you don't already know the answer to. Think several questions ahead of the last one you asked , put them into a corner by waiting for them to respond. Do not get upset , do a little homework before calling ( your reading this which means your on the right track already )

Pay close attention to the testmy.net database of ISP's and there results overall. The data stored at testmy.net is compiled over long periods of time and hundreds of ISP's around the world , which makes it a great place to do a little more digging into the past present performance of your ISP in general.

Call your ISP ( repeatedly ) and when they ask if you would like to have someone come out and test , always say yes , of course there are those that charge to have a tech come out ,  and in some instances you are left with no choice if you would like to get results. This does several thing , first and most important , it creates a record of your issues. Second , with that data already in there systems , you can use this further down the road to get to a higher level technician and actually get to the point of the issue.

Make sure to test regular , as a member of testmy.net your testing records are stored indefinitely, ( unless of course you decide to delete them ) you can use this information to show the technician your connection averages, if they claim they cannot guarantee speeds outside of there network , this is true in many ways , but in others , they have much more control then they might lead on. After all , most businesses are in it for the money , and the less time they spend on technical calls to a higher paid employee the better.

Just always be kind to customer service , it usually gets you further. Educate yourself by reading through others issues in the testmy.net forum, see what solved there issue , find out what there ISP said to them and what the outcome is / was. Asking educated questions and giving educated answers that an entry level customer service might not be well versed on , or is not at there disposal to answer , will likely get your call elevated to a 2nd or even 3rd level tech. Which in turn raises your chances of looking further into the issue and your quest , getting what you are paying for.

this is completely right, when i was having issues with my connection months back i had to call tech support so many times i knew most by name and had some technicians phone numbers, i must've called support 100 times over the course of 3 months and had countless tech's come check my connection... i got so sick of answering the typical have you tried connecting directly to the modem? have you tried a different modem?? i even went and bought 2 extra modems and got an extra modem from COX because it was their approved modem and all offered little to no change...  stay vigilant and its never a bad idea to keep a record of who you talked to.. keep names written down and notes of what they said so if you talk to someone who is unfriendly and questions you... you have hard proof

[kenwood850]

Thanks to all that responded, Mudmanc4 and TriRan for their insightful advice on how to deal with tech. support, RTB for the answer to the 90% question and of course CA3LE for not only his comments but for running this site.

I have always been aware of the fact that Ethernet had overhead but I would have expected that most of the flash speed tests sites, at least those run by ISPs, would have counted that overhead in their total speed value since it is to their advantage to show the highest speed possible.  Also RTB’s comments suggesting that value was 10% seemed a little high so I did some more poking around and found out that ADSL breaks down an Ethernet packet into cells 53 bytes long and that adds even more overhead for a total of over 16%.   I was not aware of that so I learned something new.  I guess since I really do not know what the speed figure really represents, I will accept within 10% being a satisfactory reading.

As for the location of the ISP’s speed test site, I live in the northern New England and the first two hops are local within 30 miles but then it is a direct link (at least no hops shown) to a site in TN.   That is very unusual as even traffic within my state has many more hops and all of the traffic goes south to NYC before it is routed back north – or anywhere else for that matter.

I also know were TriRan is coming from as I have talked to tech. support perhaps 6 times under 4 tickets and have had techs at my home twice and they have found nothing wrong (there generally isn’t even a problem during the workday) with my configuration.  I have also had to wipe all my NAT values, hookup direct instead of to my LAN, and use their modem instead of my more capable modem on multiple occasions with the same results I find with my own configuration.  I understand their reasoning but although it has not been the 100 times RTB has seen, it sure does get old to spend an hour each time covering old ground.

I have one more question.  Is anyone aware of how to automate retrieval of test speed readings from an Ookola flash speed test site?  I know you can automate tests with testmy.net but I would really like to do it with my ISP’s test site.   I am familiar with bash scripting so ideally it would be something that would run from a Linux (Slackware) command line that I could call using crond.  On the other hand something that worked with Windows would also be acceptable.  Beggars cannot be choosers.  

Thanks again for all your comments.

[CA3LE]

kenwood850, on 30 October 2011 - 06:14 PM, said:

...and of course CA3LE for not only his comments but for running this site.

No problem

kenwood850, on 30 October 2011 - 06:14 PM, said:

I have one more question.  Is anyone aware of how to automate retrieval of test speed readings from an Ookola flash speed test site?  I know you can automate tests with testmy.net but I would really like to do it with my ISP’s test site.   I am familiar with bash scripting so ideally it would be something that would run from a Linux (Slackware) command line that I could call using crond.  On the other hand something that worked with Windows would also be acceptable.  Beggars cannot be choosers.  

Nope, the Auto Speed Test is unique to TestMy.net.  Even with scripting on your end it's nearly impossible unless the test site wants you to be able to do that.  You'd need to be able to automate mouse clicks to get the test started otherwise.  Far too much trouble IMO.  

You could write a bash script that downloads a file from one of their servers and logs the results... sounds like allot of work considering that I've already built the tool you need.     --- Although, upcoming versions of the automatic speed test will have the option to test from any server you want.  It should be sweet but I bet your problem is resolved by the time I release that update.

Thanks for visiting!  Tell your friends...

[hoanimas]

I have just acquired a link for my office. The ISP contract is 800 down and 400 Kbps with contention ratio of 4:1. I am expecting that I will get on the minimum 200/100 Kbps. I have seven days from two days ago (27th Nov 2011) to test and accept or reject the connection service. I want to do as many test as possible. I fortunately, I found an option of Auto-Speed-Test that allow me to do multiple test in the background but I am not able to use it on my iMac. I need someone to help direct me quickly to how to set this up. I want a test every 30 mins. I also want to know if load on the link while in use by my office will affect the result of the Speed-Test.

[CA3LE]

hoanimas, on 29 November 2011 - 02:35 AM, said:

I have just acquired a link for my office. The ISP contract is 800 down and 400 Kbps with contention ratio of 4:1. I am expecting that I will get on the minimum 200/100 Kbps. I have seven days from two days ago (27th Nov 2011) to test and accept or reject the connection service. I want to do as many test as possible. I fortunately, I found an option of Auto-Speed-Test that allow me to do multiple test in the background but I am not able to use it on my iMac. I need someone to help direct me quickly to how to set this up. I want a test every 30 mins. I also want to know if load on the link while in use by my office will affect the result of the Speed-Test.

Welcome to TestMy.net hoanimas!

http://testmy.net/auto is where you need to go to get started.  Here's a video I just made for you to show you what to do.



I recommend that you refrain from using the connection during the test.  You'll get the most accurate results if you do it that way.