IDNetters Forums

Technical News & Discussion => IDNet Help => Topic started by: quandam on May 06, 2009, 19:00:59

Title: Another Silly Question For The Experts
Post by: quandam on May 06, 2009, 19:00:59
My connection is superb and when using the internet everything runs smoothly and quickly. :thumb: Clearly not a problem, and I should be very happy with events. :thumb: However, when I do the usual speed tests (varied) my speeds show up as average or below. How can this be?

I am very happy with how things pan out but how do I get reasonably poor speeds showing up on the speed tests? I was advised to (via IDNetters)  switch interleaving off and since then the speed tests have dropped dramatically.

I know I can be accused of 'what is the problem then'? but it is a puzzle why my speeds via testers (varied)  do not reflect my everyday useage experience?

Any ideas?
Title: Re: Another Silly Question For The Experts
Post by: Baz on May 06, 2009, 19:13:27
that looks ok to me

heres my result from that site you posted (http://www.mybroadbandspeed.co.uk/results/03698565.png) (http://www.mybroadbandspeed.co.uk)

Im not an expert but the speed test sites are not that reliable.What do you get using the tester on BT site and what are your router stats showing
Title: Re: Another Silly Question For The Experts
Post by: Glenn on May 06, 2009, 19:17:43
The only test BT take notice of is their own, here is the site that you used on my poor line

(http://www.mybroadbandspeed.co.uk/results/03698677.png) (http://www.mybroadbandspeed.co.uk)
Title: Re: Another Silly Question For The Experts
Post by: gizmo71 on May 06, 2009, 20:31:59
Quote from: Glenn on May 06, 2009, 19:17:43
The only test BT take notice of is their own, here is the site that you used on my poor line

(http://www.mybroadbandspeed.co.uk/results/03698677.png) (http://www.mybroadbandspeed.co.uk)

"poor"?! That's nearly twice what I'm getting on my wet piece of string! ::) :o :bawl: :laugh:
Title: Re: Another Silly Question For The Experts
Post by: Glenn on May 06, 2009, 21:02:24
Dave, I said poor, not bad  ;)
Title: Re: Another Silly Question For The Experts
Post by: David on May 06, 2009, 21:19:28
Quotewhat is the problem then

As far as I am aware Q all speed sites are correct although they may not tally up with each other I found the best results is to use one site like TB or whoever and carry out your tests over a week or how ever long you wish to torment yourself and you will get an average I think thats as close as it gets.

The problems or uncertainties arise when we compare a speed from one site with another and although the results may differ they are in all probability correct.

I don't know what I'm talking about.......but I think I m correct in this answer  :fingers:
Title: Re: Another Silly Question For The Experts
Post by: Dopamine on May 06, 2009, 23:06:53
Ever since thinkbroadband.com made available their test files for download (http://www.thinkbroadband.com/download.html) I've stopped doing speed tests and used those files instead in conjunction with Netmeter. That allows me to see a continuous real-time speed graph. Speedtests, even the best, only download a limited amount of data, whereas I can set a large download running and watch it for many minutes. Ideal for tracing intermittent faults, or wireless signal problems.
Title: Re: Another Silly Question For The Experts
Post by: Simon on May 06, 2009, 23:47:10
Sounds like good advice, Dopamine.  :)
Title: Re: Another Silly Question For The Experts
Post by: Sebby on May 07, 2009, 00:23:36
Speedtesters are notoriously inaccurate and variable. The results often have to be taken with a pinch of salt.

I've also found the Namesco (http://www.speedtest.bbmax.co.uk/) tester to be consistent and accurate.

As Dopamine suggests, downloading a file is often the best way to test your line, or I find a well-seeded torrent does the trick (such as OpenOffice).
Title: Re: Another Silly Question For The Experts
Post by: quandam on May 07, 2009, 11:22:06
Thanks All. As I said earlier I really have very little to moan about but it did seem odd that the tests didn't match up to the 'feel' in actual use. :dunno:
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 12:01:15
I think there are too many variables with speed tests, Q, your AV software can affect the results, for example. Downloading a reasonably large file is about the only reliable method, and even that's dependent on the server load at the time you do it.
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 17:18:22
The only reason I use speed tests nowadays is that they occasionally show that, on download speed at least, I am on ADSL2+.

It's somehow reassuring to know that I can actually get there. Sort of....  ;)
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 17:29:05
Have you been WBC'd then, Bob? If so, how did it go?
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 17:37:35
No....that's the point.  ;D

The speed tests are so inaccurate that they occasionally record download speeds of the ADSL2+ variety.

For example,

(http://www.speedtest.net/result/448534584.png) (http://www.speedtest.net)

But the upload always gives it away.
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 17:43:07
Ah. It's just that I'm thinking about making the move, so am looking for people who have done it.
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 18:32:05
Not available is it?
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 18:33:11
Soon. :)
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 18:38:38
Aaaahhh!  ;D

I was going to say. I go away for a few days and look what happens.  :D

I've already inquired some time ago just after I got de-interleaved. Told them to put me on the list.

Can't do any more ATM. I don't mind being a pioneer just so long as I get a faster connection.
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 18:42:20
Remind me of your line stats, will you, Bob?
Title: Re: Another Silly Question For The Experts
Post by: Sebby on May 07, 2009, 21:06:41
Quote from: bobleslie on May 07, 2009, 17:37:35
No....that's the point.  ;D

The speed tests are so inaccurate that they occasionally record download speeds of the ADSL2+ variety.

For example,

(http://www.speedtest.net/result/448534584.png) (http://www.speedtest.net)

But the upload always gives it away.

That'll be your antivirus/security suite causing that. Try turning it off then running a test. :)
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 21:23:00
Quote from: Rik on May 07, 2009, 18:42:20
Remind me of your line stats, will you, Bob?

Here's a freshly minted one, Rik:

Quote/bin/cat /proc/avalanche/avsar_modem_stats=
AR7 DSL Modem Statistics:
--------------------------------
[DSL Modem Stats]
   US Connection Rate:   448   DS Connection Rate:   8128
   DS Line Attenuation:   23   DS Margin:      11
   US Line Attenuation:   14   US Margin:      30
   US Payload :      28389648   DS Payload:      176321232
   US Superframe Cnt :   7252209   DS Superframe Cnt:   7252209
   US Transmit Power :   11   DS Transmit Power:   19
   LOS errors:      0   SEF errors:      0
   Errored Seconds:   0   Severely Err Secs:   0
   Frame mode:      3   Max Frame mode:      0
   Trained Path:      0   US Peak Cell Rate:   1056
   Trained Mode:      3   Selected Mode:      1
   ATUC Vendor Code:   414C4342   ATUC Revision:   1
   Hybrid Selected:   3   Trellis:      1
   Showtime Count:      1   DS Max Attainable Bit Rate: 10720 kbps
   BitSwap:      1   US Max Attainable Bit Rate:   n/a
   Annex:          AnxA   psd_mask_qualifier: 0x0000
   Power Management Status: L0   DS HLINSC: 0
   US ACTPSD:       -345   DS ACTPSD: -365
   Total init. errors:    0   Total init. timeouts: 0
   Showtime init. errors:    0   Showtime init. timeouts: 0
   Last showtime init. errors: 0   Last showtime init. timeouts: 0
   ATUC ghsVid:  0f 00 41 4c 43 42 00 00
   T1413Vid: 00 00      T1413Rev: 00      VendorRev: 00
   ATUR ghsVid:  b5 00 54 53 54 43 00 00
   T1413Vid: 00 00   T1413Rev: 00   VendorRev: 00

   [Upstream (TX) Interleave path]
   CRC:    0   FEC:    0   NCD:    1
   LCD:    0   HEC:    0

   [Downstream (RX) Interleave path]
   CRC:    0   FEC:    0   NCD:    0
   LCD:    0   HEC:    0

   [Upstream (TX) Fast path]
   CRC:    1   FEC:    0   NCD:    0
   LCD:    0   HEC:    0

   [Dow
stream (RX) Fast path]
   CRC:    85   FEC:    0   NCD:    0
   LCD:    0   HEC:    0

[ATM Stats]
   [Upstream/TX]
   Good Cell Cnt:   591451
   Idle Cell Cnt:   129674642

   Tx Packets Dropped Count:   0
   Tx Bad Packets Count:   9

   [Downstream/RX)]
   Good Cell Cnt:   3673359
   Idle Cell Cnt:   2359725648
   Bad Hec Cell Cnt:   108
   Overflow Dropped Cell Cnt:   0
   Rx Packets Dropped Count:   0
   Rx Bad Packets Count:   0


[SAR AAL5 Stats]
   Tx PDU's:   118742
   Rx PDU's:   176412
   Tx Total Bytes:   23808401
   Rx Total Bytes:   171832189
   Tx Total Error Counts:   0
   Rx Total Error Counts:   1


[OAM Stats]
   Near End F5 Loop Back Count:   0
   Near End F4 Loop Back Count:   0
   Far End F5 Loop Back Count:   0
   Far End F4 Loop Back Count:   0
   SAR OAM Ping Response Drop Count=0

It's a 7150 profile under ADSL.
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 07, 2009, 21:31:43
Quote from: Sebby on May 07, 2009, 21:06:41
That'll be your antivirus/security suite causing that. Try turning it off then running a test. :)

Thanks. I usually just do another test.

It doesn't happen all the time, but it makes for a great conversation piece.  :D

On a 'normal' day,  I reckon to get 6700 - 6900kbps, which is what I would expect for my 7150 profile. Remarkably this has been extremely consistent on the two separate occasions I've been with IDNet.
Title: Re: Another Silly Question For The Experts
Post by: Sebby on May 07, 2009, 21:32:22
That's what we like to hear. :)
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 07, 2009, 22:20:19
You're in line for very good speeds, Bob, 20-24M I would say.
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 08, 2009, 11:13:40
Hope you're right, Rik.

I'd be happy with that. :whistle:
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 08, 2009, 11:35:03
Now I've got access to the calculator, I can give you a better answer, sadly not quite as good as I first thought. Sync ~19M, throughput ~ 16.8M.
Title: Re: Another Silly Question For The Experts
Post by: bobleslie on May 08, 2009, 12:28:55
That sounds more like it, Rik.

I'd still be happy at those figures.  ;)
Title: Re: Another Silly Question For The Experts
Post by: Rik on May 08, 2009, 12:42:20
So would I. The best I'll see is a 500k gain I think.