BT speedtester

I've just run a BT speedtest, rubbish result as usual, but the list of items you now have to say "Yes" to before you can proceed is interesting:

Step 1:- Ensure your computer is plugged into the Broadband line to be tested.This speedtester will not work on iPads/Tablets,smartphones or other mobile devices.
Step 2:- Ensure your computer is directly connected to your modem/router via an Ethernet Connection (i.e. not via a wireless or powerline adapters).
Step 3:- Ensure that any wireless adaptor in your computer is switched off.
Step 4:- Close any programs that may be running on your computer.This in cludes any background programs such as anti-virus software (such as Norton McAfee, or NetProtect+) corporate VPN or Peer2Peer client.(Please ensure you re enable any anti-virus software after you have finished testing).
Step 5:-Reboot your modem/router by powering it off,waiting one minute,and then powering it up again.Wait for any lights on the modem router to stabilise before the starting the test.
Step 6:-Restart your browser.
Step 7:-Ensure that no other people or devices(eg. broadband-connecterd TV set-top box) are using the broadband line to be tested.
Step 8:-Ensure the 10 day stabilisation period has completed before reporting any suspected fault-this is how long it can take for your Broadband to train to its optimal speed.
Step 9:-For fibre customers please run the Beta test first (right-hand column below),ensuring you use the correct test for your line speed.

Please confirm that you have carried out the above steps. The speed tester may report an inaccurate result by failure to comply with the above.

Note number 8... the words "left and right hand" come to mind

I noticed the new wording earlier this evening. Wondering if this has has anything to do with the inconsistent and inaccurate results it has been giving recently and in particular the FTTC results.

See they are also recommending that you use the Beta tester for FTTC/P connections.

Edited by nelix01 (Thu 26-Apr-12 20:40:55)

It seems more to me¹ as though they're giving themselves as many reasons as they can think of to blame the user for poor performance.

Most of the things on there won't make the slightest difference, let alone a result of 2Mbps on an 80/20 line that I've just had



¹ Cynic? Moi? Naah...

there just trying to make it more difficult for end users to complain about speeds. basically buying time fore the poor network

I've had 80/20 for almost 3 months now and never had a decent test result using the vanilla tester. The Beta tester almost always gives me a result of 75Mbps.

Same here, I really don't see the point of the BT tester in its current form as it is totally incapable of giving sensible results; I get anything from 0 upwards on it and seems completely arbitrary. Yet this is the "official" tester.

Yes, the beta works OK for me, but it doesn't tell you what your IP profile is- I'm usually more interested to see what might have happened to that than in the speed.

I suppose it's the first step to blocking access to profile information

Edited by billford (Thu 26-Apr-12 21:05:41)

So in non-BT speak...

4. Please stop your anti-virus software (who needs it anyway, there aren't that many Chinese teenagers looking at hacking your ebay account. It'll be perfectly safe switching it off - we're BT, you can trust what we say).
5. Please reboot your Home Hub (as we all know how flaky they are).

They forgot 9,10 & 11...

9. Please power down the cheap Chinese modem and let it cool down for a few hours.
10. Please do a beta test to get your download speed (because the normal one is rubbish with FTTC).
11. Please do a normal test to get your upload speed (because the beta test is rubbish with FTTC uploads and often returns a figure around 40% of your actual speed).

Oh and 12...

12. Please forgot about speedtesters, they're all rubbish anyway. Just do a multi-thread download (using a download manager program) if you really want to know your download speed.

In reply to a post by billford:

Yes, the beta works OK for me,

Clearly the beta test isn't subject to whatever throttling and shaping BT are imposing at the moment (as it's giving a true download result of 74Mbps when all the other speedtesters get around 4-8Mbps).

In reply to a post by billford:

but it doesn't tell you what your IP profile is- I'm usually more interested to see what might have happened to that than in the speed.

I'm guessing it's just a regular java based speedtest (but one which bypasses BT throttling) and they haven't yet bothered to link it to the number database to give the profile figure (hoping they'll add it if it ever goes beyond beta - which would be gamma; no?).

I've not had any speed issues so far but have unlocked my spare modem today so I can keep an eye on the sync rates.

My line is due to go 80/20 on the 9th, but looking at the stats it will be more like 54/17 and probably start dropping as the cabinet starts to fill up.

Lets hope I never have to use the BT Speedtester. A pilots pre-flight checklist is probably less traumatic!

In reply to a post by adebov:

if it ever goes beyond beta - which would be gamma; no?

Well, gamma rays are a lot more penetrating than beta, but they can also do a lot more damage

Step 8:-Ensure the 10 day stabilisation period has completed before reporting any suspected fault-this is how long it can take for your Broadband to train to its optimal speed.

[weep]

How wrong can they be? They are wrong about the 10-day period if on ADSLx, and completely up the spout if it is FTTx. Then we have

Step 9:-For fibre customers please run the Beta test first (right-hand column below),ensuring you use the correct test for your line speed.

For a quick and easy test of your broadband speed(takes about 30 secs) please click on the relevant Beta test button below.For a further diagnostic check (takes upto 3 mins),please click on the relevant 'Diagnostic' test.

They have just laid the page out appallingly. There should be a third note holding both beta tests. The current layout is self-contradictory as it effectively applies Note 1 to the up to 24Mbps and Note 2 to the over 24Mbps.

The mods on the bt forum are adamant about the 10 day training period now aswell, they claim to have double checked it's existance and that most faults aren't accepted within the first 10 days:-

http://community.bt.com/t5/BT-Infinity/Bt-Infinity-u...
http://community.bt.com/t5/BT-Infinity/Speed-dropped...

I think this "10 day stabilisation period" claim is being used to allow time for any congestion to be alleviated. I've seen a few instances of people reporting slow speeds on the bt forums which have sorted themselves out by the end of the 10 days.

Well it doesn't exist in either SIN 498 which is Openreach's description of it, in particular Section 2.2.5

Dynamic Line Management (DLM) is employed in GEA-FTTC. DLM constantly manages lines to maintain a target stability. It does this for as long as the product exists.
At provision, the line is put on wide open profiles, allowing downstream line speeds of up to 40Mbit/s, and upstream line speeds of up to 2Mbit/s, or 10Mbit/s. depending on the upstream product option selected.
On the first day of operation, DLM will intervene if severe instability is detected. Otherwise, DLM will wait until the day after provision before intervening, provided that the line has been trained up for at least 15 minutes during the preceding day.
If DLM intervenes it will set a capped profile with a maximum rate and a minimum rate, where the minimum rate is set at approximately half of the maximum rate.

nor in SIN 495 which is BT Wholesale's description of their part of it.

That refers to SIN 498 and also to SIN 472, the standard WBC document, with this note

This SIN (472) provides the interface information between BTW�s WBC product and the SP. All sections are relevant with the exception of Section 3 End User Access.

As this has not been revised since Issue 1 in November 2009, I suggest it should also exclude the section about WBC DLM, or at least the 10-day period for line stabilisation. The line stabilisation is under the control of the OR DLM in the cabinet, as above.

It even says

The main difference between the WBC FTTC service and the WBC ADSL1 and 2+ Service is that the DSLAM (�Digital Subscriber Line Access Module�) for the WBC ADSL1 and 2+ Service is situated in an exchange building but for WBC FTTC the DSLAM is situated in a street cabinet (the �Street DSLAM�).

Joined up thinking was not present in 2009, when I don't think the system was only in its infancy. Or was it only in development?

Whichever, the stabilisation by the OR DLM supercedes and replaces any possible stabilisation by the BTW one. The only possible relevance of the BTW DLM logically has to be to forward the pre-existing ISP-requested stability settings to the OR one, i.e. the Standard, Stable and Super Stable options, and possibly to set the IP Profile, thus saving completely re-engineering those parts of the system.

Note - I haven't yet found any SIN that has been updated for 80/20.
Edit - it has been brought to my attention that the link to SIN 472 went to SIN 495. Now corrected.

Edited by RobertoS (Fri 27-Apr-12 13:20:20)

FWIW, my ISP told me that yes, the 2 day DLM "training" period is indeed valid, but up to 10 days have to be allowed for the connection's MSR (Maximum Stable Rate) to be determined before "faults" can be reported.

In reply to a post by Bald_Eagle1:

up to 10 days have to be allowed for the connection's MSR (Maximum Stable Rate) to be determined before "faults" can be reported.

Then they should say so-

"Note that we require a 10-day period to determine the normal range of speeds on your line before we can accept a low speed as a fault"

is purely administrative matter and totally unrelated to:

"... this is how long it can take for your Broadband to train to its optimal speed."

which is plain simple wrong. The line has "trained" to its optimal speed as soon as the modem and DSLAM finish their negotiations and, as Bob said, this continues for as long as the line is active.

eta- Under the rule as stated, if I were a new user then the sub-5Mbps tests I've been getting for a month or more on the 80/20 line would not be accepted as a fault (even though the fault is clearly with the tester itself, not the line).

Edited by billford (Fri 27-Apr-12 09:24:55)

I've given up with them.

Potentially making people wait for 10 days, could be seen as way to circumvent Distance Selling Regulations, i.e. so that people cannot get out of service early if it performs well under the speeds.

In reply to a post by MrSaffron:

I've given up with them.

I can understand that.

Use the words "talking to", "BT", "banging", "head" and "brick wall" in a apposite phrase or sentence

My day rate to re-write the page for them is fairly cheap these days

I'm guessing it's just a regular java based speedtest (but one which bypasses BT throttling) and they haven't yet bothered to link it to the number database to give the profile figure (hoping they'll add it if it ever goes beyond beta - which would be gamma; no?).


Nope in software development terms the gamma build is the first in house one where the code is put together and sees if it even works.

The next step after beta is the alpha or release version.

In reply to a post by Anonymous:

The next step after beta is the alpha or release version.

Cobblers.

Alpha => Beta => (Public Beta) => Release candidate => Final version.

Until the Service pack to fix the undiscovered bugs, anyway

BT speedtester is a load of tosh any way, never really been able to get a proper result off it, Just done 1 and it's come back as:-
Download speed achieved during the test was - 1Mbps IP Profile 77.43Mbps
Upload speed achieved during the test was - 16.38Mpbs IP Profile 20Mbps
The part I like is:-
We are unable to identify any performance problem with your services at this time.
Classic.
iechyd da

In reply to a post by paul1360:

BT speedtester is a load of tosh any way

Agreed, but afaik it's the only one they'll take any notice of

That's the silly part about it ISP's will only take notice and yet what would they say if I rang up would they look into it or take no performance problem's. lol crazy
iechyd da

In reply to a post by MrSaffron:

My day rate to re-write the page for them is fairly cheap these days

No doubt, but then BT would have to address customers' legitimate grievances or even admit that they might be at fault... and we can't have that

Edited by billford (Fri 27-Apr-12 11:25:01)

In reply to a post by Croftie:

The mods on the bt forum are adamant about the 10 day training period now aswell, they claim to have double checked it's existance and that most faults aren't accepted within the first 10 days:-

http://community.bt.com/t5/BT-Infinity/Bt-Infinity-u...
http://community.bt.com/t5/BT-Infinity/Speed-dropped...

I think this "10 day stabilisation period" claim is being used to allow time for any congestion to be alleviated. I've seen a few instances of people reporting slow speeds on the bt forums which have sorted themselves out by the end of the 10 days.

For what it's worth my line saw a drop in latency yesterday morning:

My Broadband Ping

The interesting thing is that my connection went live Monday the 16th and it was mid-morning before the engineer had finished faffing around. So that's ten days quite possibly to the hour. Now perhaps it's a coincidence but seems unlikely. Oh and as noted in another thread my upstream speed on TBB jumped up about 1Mb/s. Didn't change on the other testers I use but then they always reported it as higher than TBB. It now nearly brings TBB into the same ballpark.

Edited by Andrue (Fri 27-Apr-12 11:43:07)

In reply to a post by Bald_Eagle1:

FWIW, my ISP told me that yes, the 2 day DLM "training" period is indeed valid, but up to 10 days have to be allowed for the connection's MSR (Maximum Stable Rate) to be determined before "faults" can be reported.

First, your ISP has no choice but to repeat what brainwashed BT staff tell them. It does not make the BT statement correct. The BT documentation itself debunks the myth.

Second, in the case of ADSLx, it is true that the 10-day period establishes the MSR and hence the FTR. Nobody has ever disputed that. But there are many cases where there is a clear fault during that period, where an extremely low sync can occur, not just a feasible DLM line adjustment.

In such cases the fault needs to be corrected and the 10 days re-started. Otherwise the MSR and FTR are incorrect and the whole point of the 10-day period is negated. Hah - fat chance! (It does occasionally happen, but I think only with on-the-ball ISPs).

Third, in the case of FTTC, the MSR and FTR may be retained by BT Wholesale in their DLM, for ease of implementation similar to those described in my previous post, but they are totally irrelevant to the Openreach process line management system. The initial stabilisation takes place in the first couple of days, and that's all there is to it. If BT Wholesale continue, for some unkown reason, to want to wait 10 days before setting the MSR/FTR for admin purposes, then they should change the name to something like Line Assessment Period. Even in the two community threads linked to earlier they still say "stabilisation period". It simply, incontrovertibly. has nothing to do with stabilising anything.

In fact it contributes greatly to the destabilisation of customers' confidence in BT.

In reply to a post by RobertoS:

In fact it contributes greatly to the destabilisation of customers' confidence in BT.

I couldn't agree more.

It seems like BT OR should by now have issued a FTTC/VDSL2 "Myths & Legends" document to help us numpties actually understand (see through) the drivel that some CS agents read from a script.

BT OR should by now have issued a FTTC/VDSL2 "Myths & Legends" document

Wow hang on this is BT we are talking about and to do that would be to easy for them and that's 1 thing BT don't do "Easy", that's like Sky letting all android devices use sky go.
lol not going to happen
iechyd da

In reply to a post by Anonymous:

Nope in software development terms the gamma build is the first in house one where the code is put together and sees if it even works.

The next step after beta is the alpha or release version.

Did not know that. I always assumed 'alpha' was a pre-beta/internal release and my gamma comment was a poor attempt at humour.

In reply to a post by adebov:

I always assumed 'alpha' was a pre-beta/internal release

It is.

No wonder he/she posted as anon

Some Anons are less anonymous than others

Edited by billford (Fri 27-Apr-12 20:14:49)

I just got pretty much the same speed test result too Bill 2MB on the 80/20 HA HA this BT Tester is a joke.

Indeed, what a joke:- http://imgbox.com/aae6upA6 - 2AM

Looking at my bandwidth monitor It didn't actually manage to transfer anything on the download, so no supprise it reported 0Mb.

Seems to be in a better mood ATM: http://imgbox.com/aagUDvDn - 11PM

In reply to a post by adebov:

So in non-BT speak...

4. Please stop your anti-virus software (who needs it anyway, there aren't that many Chinese teenagers looking at hacking your ebay account. It'll be perfectly safe switching it off - we're BT, you can trust what we say).
5. Please reboot your Home Hub (as we all know how flaky they are).

They forgot 9,10 & 11...

9. Please power down the cheap Chinese modem and let it cool down for a few hours.
10. Please do a beta test to get your download speed (because the normal one is rubbish with FTTC).
11. Please do a normal test to get your upload speed (because the beta test is rubbish with FTTC uploads and often returns a figure around 40% of your actual speed).

Oh and 12...

12. Please forgot about speedtesters, they're all rubbish anyway. Just do a multi-thread download (using a download manager program) if you really want to know your download speed.

LOL couldnt agree more the only reliable speedtest ive found is http://vmspeed.com/

In reply to a post by Bapteeed:

the only reliable speedtest ive found is http://vmspeed.com/

Er . . don't think so. I'm at the end of a long copper wire and am currently sync'd at 2624kbps which means a BT IP profile of 2Mb/s. The Virgin test (as with all of the Ookla based tests) gets optimistic on lines like that and has just returned me a download speed of 2.07Mb/s which is (in theory) impossible. The TBB speed test has just returned 1.8Mbs which seems a lot more in keeping with the line.

Yes...DLM intervenes it will set a capped profile with a maximum rate and a minimum rate.