Switched to Infinity 2 this morning

engineer says it's sync'd up at 80 / 20

The BTO engineer was done by 9am

Predicted (by the online checker) was 65.3 down and 20 up

[IMG]http://www.speedtest.net/result/1929586172.png[/IMG]
74 down 16 up

The tester on here (TBB) shows results that are lower than expected.
(hence not posted the result)


just need to change my sig on here now.

Edited by richarddawson (Thu 03-May-12 12:36:09)

In reply to a post by richarddawson:

engineer says it's sync'd up at 80 / 20

The BTO engineer was done by 9am

Predicted (by the online checker) was 65.3 down and 20 up

[IMG]http://www.speedtest.net/result/1929586172.png[/IMG]
74 down 16 up

The tester on here (TBB) shows results that are lower than expected.
(hence not posted the result)


just need to change my sig on here now.

Oooh, nice ping, too.

Enjoy it!

That sorts out the question raised in http://forums.thinkbroadband.com/bt/t/4119412-infini...

A customer will get the max the sync speed will allow and it will NOT be restricted by the checker predictions..

He was predicted 20 upstream and is getting 16. Doesn't really disprove anything.

It told me 20Mb up but ended up with 15Mb then after several days it picked up to 18Mb and has been stable ever since.
iechyd da

PS
No matter what I do or try I can't seem to change my sign, any idea's as to why?

Edited by paul1360 (Fri 04-May-12 09:45:33)

In reply to a post by paul1360:

No matter what I do or try I can't seem to change my sign, any idea's as to why?

Astrology is based on your date of birth, which is fairly constant.

Assuming you are going to the right place to change your sig, are you forgetting to click the Submit button at the bottom?

Remember that the observed throughput will always be lower than the sync rate.

On ADSL IIRC the maximum expected throughput was about 85% of sync rate. I'm told the expected maximum throughput on FTTC is more like 92%, though my results seem to agree with the 85: When I was on 40/10 the best speedtest results where 8.5mbit upstream IIRC (as far as I know I was syncing at the full 10 but I've not unlocked the modem to find out for sure), I've just moved to 80/20 and I'm now seeing 16.5 as the best result upstream though I've not done much testing yet.

And to think that I remember when you had to pay a hefty price to push any part of your local wired network faster than 10mbit/s...

My experience is that "observed" downstream speeds will be 97% of your sync thanks to the IP profile, and user-data throughput (i.e. your average HTTP speedtest result) will be about 94% of your sync thanks to an additional 2-3% of TCP overheads.

On the upstream however observed speeds are closer to 97% because there is no IP profile (in practice) - but most speed testers generally fail at testing fast upstream speeds.

Speedtest.net for example, usually comes up with 16.5mbps on mine line, TBB's speedtest shows only 6mbps, but Visualware's UDP test shows 18.5mbps. However a single FTP upload to my own server proceeds at 19.4mbps.

In any case the quick investigation of Speedtest.net's results shows bursty transmission and my line sitting idle for a third of the test, and a result that was considerably higher than the actual speed of data sent on the line during the test.

Basically, most speedtests are especially [censored] at uploads.

Edited by qasdfdsaq (Fri 04-May-12 14:30:50)

In reply to a post by qasdfdsaq:

... but most speed testers generally fail at testing fast upstream speeds.

Speedtest.net for example, usually comes up with 16.5mbps on mine line, TBB's speedtest shows only 6mbps ....

How close are your tbb upstream results to these? :-

02/02/12 00:49 Thu 36401 Kbps 6068 Kbps
31/01/12 17:06 Tue 35457 Kbps 6061 Kbps
22/01/12 10:59 Sun 36388 Kbps 6077 Kbps
07/12/11 00:24 Wed 35688 Kbps 6007 Kbps
18/11/11 23:57 Fri 36249 Kbps 5999 Kbps

In reply to a post by richarddawson:

The tester on here (TBB) shows results that are lower than expected.
(hence not posted the result)

Please do, it might help the guys at TBB work out why the TBB tester is so flaky with BT Infinity2 connections.
Here's mine (which is, BTW, the best result I've had from the TBB tester in weeks).


Here's my Speedtest.net result (for comparison).

It is worth pointing out if you try several of the Speedtest.net servers, you'll find a few in the UK which give similar results to the TBB tester.
I doubt if some Speedtest.net servers use multiple thread tests, while others use single thread tests (which has been touted as one possible explanation why the TBB test results are so much lower) so it points to some other problem (maybe some servers are being overloaded - possibly one being the TBB test server or BT have some routing issues).

If I try a Speedtest.net with two Xilo servers (one in Manchester, the other in Maidenhead) I get vastly different results (around 75Mbps/16Mbps and a ping of 10ms for the Manchester server, and around 28Mbps/12Mbps and a ping of 26ms for the Maidenhead server).
I'd find it hard to believe one Xilo server is applying the test in a vastly different way from another Xilo server, so it has to be to do with BT (throttling???), routing, or server congestion.
This makes no sense to me (unless BT are applying some far out routing) as I'm 44 miles from Maidenhead and 107 miles from Manchester (direct "as the crow flies" distances), so God knows where BT are routing my connection (if the ping to do a 214 mile round trip is 2.6 times lower than the ping to do an 88 mile round trip).

Edited by adebov (Fri 04-May-12 17:41:31)