Last time I ever used IE
Download speeds
Moderator: Moderators
-
trainlover123
- Very Active Forum Member
- Posts: 1334
- Joined: Mon Feb 09, 2004 9:27 pm
- Location: Doing something worthless
-
richard222
- Very Active Forum Member
- Posts: 1572
- Joined: Tue Oct 10, 2006 2:44 pm
- Location: Surrey
- Contact:
-
Easilyconfused
- Worried about Silent Chickens
- Posts: 13205
- Joined: Tue Dec 31, 2002 9:06 am
- Location: Portsmouth & Bristol
- Contact:
Hi Jon,
You are entirely correct about the ADSL speeds. I never seem to get the full speed from either of my ADSL connections. They are better during the day since I guess everyone is at work.
However, on my cable broadband I get more or less the full line speed of 4MB most of the time from a variety of sites. Now I did see a slow down a couple of weeks ago but it is back to normal now - investigations revealed Telewest doing engineering work. This is what I get most of the time :
I guess it is all dependent on where your ISP gets it's link to the USA from since I understand the servers for UKTS are in the USA. Comparing UKTS download speeds to downloads from European sites is not necessarily comparing like with like since the traffic has to go over the Atlantic rather than just to mainland Europe - gross simplification I know but I am sure Matt can fill in some more details.
Cheers
John
You are entirely correct about the ADSL speeds. I never seem to get the full speed from either of my ADSL connections. They are better during the day since I guess everyone is at work.
However, on my cable broadband I get more or less the full line speed of 4MB most of the time from a variety of sites. Now I did see a slow down a couple of weeks ago but it is back to normal now - investigations revealed Telewest doing engineering work. This is what I get most of the time :
I guess it is all dependent on where your ISP gets it's link to the USA from since I understand the servers for UKTS are in the USA. Comparing UKTS download speeds to downloads from European sites is not necessarily comparing like with like since the traffic has to go over the Atlantic rather than just to mainland Europe - gross simplification I know but I am sure Matt can fill in some more details.
Cheers
John
-
jbilton
- Very Active Forum Member
- Posts: 19267
- Joined: Fri Oct 10, 2003 12:08 pm
- Location: At home ..waiting to go to Work.
- Contact:
Hi John
Thanks....yes sounds plausible.
I would be interested to know what Glyn gets as I believe he's on 8Mb cable (he should get 750kb/s +).
Tonight for me from UKTS its a very, very low 40Kb/s....and a low but more acceptable 150kb/s from Europe.
Cheers
Jon
Thanks....yes sounds plausible.
I would be interested to know what Glyn gets as I believe he's on 8Mb cable (he should get 750kb/s +).
Tonight for me from UKTS its a very, very low 40Kb/s....and a low but more acceptable 150kb/s from Europe.
Cheers
Jon
------------------------Supporting whats good in the British community------------------------
-
NeutronIC
- Atomic Systems Team
- Posts: 11085
- Joined: Fri Oct 05, 2001 12:00 am
- Location: E11, London, England
- Contact:
The full speed is indeed available, however there are lots of reasons why it wouldn't be able to make use of it.
All I guarantee is that the code at our end will not impose any limits.
I'm on NTL 10mb connection, on a single download connection i'll generally get similar numbers to others, sometimes higher, sometimes max, but on average around the 350-450 mark also.
However, as you see above, with multiple segmented downloads via a tool such as GetRight (as i've used) much faster speeds are possible, my download above actually maxed at about 1.03mb/s in the end).
When I last upgraded the download server I did increase it's internal buffer size to see if that helped download performance - but it hasn't so I really can't see what else it would be from our end, and i'd just put it down to network issues - not necessarily your side or mine, but maybe somewhere in the middle.
In any case, I would always strongly recommend multiple segmented downloading, it's faster (that's why they call it acceleration) because one stream could get stalled or hit by something and slow it down, the chances of two suffering that are small - even less for more than two, plus they can both be streaming, even on the most optimal connection possible, two simultaneous streams will offer a very marginal (Sub 1%) performance improvement due to the way tcp/ip works, so even in the most ideal case it's not a bad thing to have, but in the normal case it's great.
I've downloaded from a wide range of sites since going to 10mb and there are very few that can fill the line unless I use multiple segmented downloads and when I do that, there are very few that *cant* fill the line.
Not saying I won't look in to it, just that I don't have a problem.
If you don't want to spend the $25 for GetRight (or whatever it costs) then there are a few freebies out there that do the same thing.
And no, I don't care how many simultaneous download connections premium users make use of for downloading, I engineered a special high performance download server specifically to allow that without sucking resources up like conventional web servers do - and when I tested upon moving to the new servers a while back, it downloaded files faster than the regular web servers (IIS and Apache) could too).
My personal settings in GetRight are set to utilise 1 connection up to a meg, 2 connections to 5 meg, 3 connections to 10 meg and 4 above that, other sites may get upset at that kind of thing but the way our download server works it won't bat an eyelid so go for it.
Matt.
All I guarantee is that the code at our end will not impose any limits.
I'm on NTL 10mb connection, on a single download connection i'll generally get similar numbers to others, sometimes higher, sometimes max, but on average around the 350-450 mark also.
However, as you see above, with multiple segmented downloads via a tool such as GetRight (as i've used) much faster speeds are possible, my download above actually maxed at about 1.03mb/s in the end).
When I last upgraded the download server I did increase it's internal buffer size to see if that helped download performance - but it hasn't so I really can't see what else it would be from our end, and i'd just put it down to network issues - not necessarily your side or mine, but maybe somewhere in the middle.
In any case, I would always strongly recommend multiple segmented downloading, it's faster (that's why they call it acceleration) because one stream could get stalled or hit by something and slow it down, the chances of two suffering that are small - even less for more than two, plus they can both be streaming, even on the most optimal connection possible, two simultaneous streams will offer a very marginal (Sub 1%) performance improvement due to the way tcp/ip works, so even in the most ideal case it's not a bad thing to have, but in the normal case it's great.
I've downloaded from a wide range of sites since going to 10mb and there are very few that can fill the line unless I use multiple segmented downloads and when I do that, there are very few that *cant* fill the line.
Not saying I won't look in to it, just that I don't have a problem.
If you don't want to spend the $25 for GetRight (or whatever it costs) then there are a few freebies out there that do the same thing.
And no, I don't care how many simultaneous download connections premium users make use of for downloading, I engineered a special high performance download server specifically to allow that without sucking resources up like conventional web servers do - and when I tested upon moving to the new servers a while back, it downloaded files faster than the regular web servers (IIS and Apache) could too).
My personal settings in GetRight are set to utilise 1 connection up to a meg, 2 connections to 5 meg, 3 connections to 10 meg and 4 above that, other sites may get upset at that kind of thing but the way our download server works it won't bat an eyelid so go for it.
Matt.
-
jbilton
- Very Active Forum Member
- Posts: 19267
- Joined: Fri Oct 10, 2003 12:08 pm
- Location: At home ..waiting to go to Work.
- Contact:
Hi Matt
Thanks for that.
I used to use Getright many moons ago...mainly to stitch downloads together...stopped using it since going onto broadband.
It looks as though your able to get a full 976kb/s....so not sure why I only ever get about 350kb/s max.....can't really be John's Atlantic link explanation.
Tonight is particularly slow
As you can see Bernard's site is nearly 3 times as fast.
Cheers
Jon
Thanks for that.
I used to use Getright many moons ago...mainly to stitch downloads together...stopped using it since going onto broadband.
It looks as though your able to get a full 976kb/s....so not sure why I only ever get about 350kb/s max.....can't really be John's Atlantic link explanation.
Tonight is particularly slow
As you can see Bernard's site is nearly 3 times as fast.
Cheers
Jon
-
NeutronIC
- Atomic Systems Team
- Posts: 11085
- Joined: Fri Oct 05, 2001 12:00 am
- Location: E11, London, England
- Contact:
Even so it looks like there's a problem since both sites are running slower than usual, it just affects our site less than it does the other, which will be related to the nature of whatever the problem is.
Remember you're not just point to point from your machine to the server, there are a dozen or more links and points between here and there, some of which may be similar to the route to bernards.
Take this for a very rough quick example:
E is you.
B is us.
A is another site you're using.
C is a different site.
D is another user.
Looking at the hypothetical routing layout, let's assume C is hammering C with perhaps a high performance link. Your access to A will be less hit by it since it only shares one hop than it would be to access B since that would incur two hops of 'traffic'.
Seems to me that everything has slowed down for you - yes, we seem to be slow for you but so does the other site, which would tend to suggest a problem closer to your end - since I doubt us and ASW share any routing outside of your ISP particularly.
Can you do a traceroute? Start / Run / Cmd and then type:
tracert http://www.uktrainsim.com
and then the same for the other site.
Paste the results and let's see where they are going.
Would also be useful if you could try a splitter like GetRight eval and see whether two connections still only get 50k/s or whether two get you 100k/s.
Matt.
Remember you're not just point to point from your machine to the server, there are a dozen or more links and points between here and there, some of which may be similar to the route to bernards.
Take this for a very rough quick example:
E is you.
B is us.
A is another site you're using.
C is a different site.
D is another user.
Looking at the hypothetical routing layout, let's assume C is hammering C with perhaps a high performance link. Your access to A will be less hit by it since it only shares one hop than it would be to access B since that would incur two hops of 'traffic'.
Seems to me that everything has slowed down for you - yes, we seem to be slow for you but so does the other site, which would tend to suggest a problem closer to your end - since I doubt us and ASW share any routing outside of your ISP particularly.
Can you do a traceroute? Start / Run / Cmd and then type:
tracert http://www.uktrainsim.com
and then the same for the other site.
Paste the results and let's see where they are going.
Would also be useful if you could try a splitter like GetRight eval and see whether two connections still only get 50k/s or whether two get you 100k/s.
Matt.
This
This anyhelp to you Matt im on BT 8mb
MARTIN
Tracing route to http://www.uktrainsim.com [209.85.59.35]
over a maximum of 30 hops:
1 31 ms 29 ms 29 ms 217.47.79.250
2 30 ms 28 ms 28 ms 217.47.86.161
3 31 ms 28 ms 29 ms 217.41.172.29
4 29 ms 29 ms 29 ms 217.41.172.65
5 31 ms 33 ms 30 ms 217.41.172.138
6 29 ms 28 ms 27 ms 217.41.216.42
7 29 ms 29 ms 28 ms 217.41.221.58
8 29 ms 28 ms 31 ms core2-pos9-1.birmingham.ukcore.bt.net [217.32.17
0.21]
9 32 ms 31 ms 32 ms core2-pos5-1.ealing.ukcore.bt.net [62.6.204.222]
10 31 ms 31 ms 31 ms transit1-pos4-0.ealing.ukcore.bt.net [194.72.9.2
38]
11 32 ms 31 ms 31 ms t2c1-p1-0.uk-eal.eu.bt.net [166.49.168.13]
12 33 ms 32 ms 32 ms t2c2-p8-0.uk-lon1.eu.bt.net [166.49.208.110]
13 33 ms 32 ms 34 ms t2a1-pc2.uk-lon1.eu.bt.net [166.49.135.106]
14 33 ms 31 ms 32 ms ge-1.linx.londen03.uk.bb.verio.net [195.66.226.1
38]
15 34 ms 32 ms 33 ms xe-0-2-0.r22.londen03.uk.bb.gin.ntt.net [129.250
.2.65]
16 109 ms 106 ms 106 ms p64-1-0-0.r21.nycmny01.us.bb.gin.ntt.net [129.25
0.2.38]
17 107 ms 106 ms 106 ms ae-0.r20.nycmny01.us.bb.gin.ntt.net [129.250.2.2
5]
18 134 ms 135 ms 134 ms p64-0-3-0.r21.chcgil09.us.bb.gin.ntt.net [129.25
0.5.4]
19 148 ms 148 ms 148 ms p64-2-2-0.r21.dllstx09.us.bb.gin.ntt.net [129.25
0.2.22]
20 152 ms 146 ms 152 ms p64-1-1-0.r21.hstntx01.us.bb.gin.ntt.net [129.25
0.4.69]
21 151 ms 154 ms 152 ms p64-1-0-0.r20.hstntx01.us.bb.gin.ntt.net [129.25
0.4.71]
22 151 ms 150 ms 150 ms xe-4-1.r03.hstntx01.us.bb.gin.ntt.net [129.250.2
.229]
23 155 ms 155 ms 154 ms ge-9.ev1.hstntx01.us.bb.gin.ntt.net [129.250.11.
134]
24 154 ms 154 ms 153 ms gphou2-209-85-1-108.ev1.net [209.85.1.108]
25 156 ms 155 ms 156 ms ev1s-209-85-59-35.ev1servers.net [209.85.59.35]
Trace complete.
-
ianmacmillan
- Very Active Forum Member
- Posts: 9588
- Joined: Fri Feb 28, 2003 12:39 pm
- Location: N. Lanarkshire Scotland
Almost doubled, but still a slow 76.4KB/sNeutronIC wrote:I've just restarted the download server to see if there was anything there slowing it down, let me know if it suddenly boosts.
Matt.
<IMG width="778" height="196" SRC="http://www.atomic-album.com/showPic.php/45092/speed.jpg">
Tracing route to http://www.uktrainsim.com [209.85.59.35]
over a maximum of 30 hops:
1 * * * Request timed out.
2 22 ms 21 ms 24 ms l1.ar02.gs1.dsl.pipex.net [62.241.161.244]
3 21 ms 21 ms 27 ms ge-0-1-0-0.cr01.gs1.dsl.pipex.net [62.241.161.10
6]
4 21 ms 21 ms 21 ms GigabitEthernet4-0.GW4.LND2.ALTER.NET [146.188.5
3.221]
5 21 ms 21 ms 21 ms so-0-0-0.xr2.lnd2.alter.net [158.43.233.118]
6 21 ms 21 ms 24 ms so-1-1-0.TR1.LND2.ALTER.NET [146.188.7.230]
7 103 ms 105 ms 105 ms so-6-0-0.IR2.DCA4.ALTER.NET [146.188.8.173]
8 105 ms 123 ms 105 ms 0.so-0-0-0.IL2.DCA6.ALTER.NET [146.188.13.41]
9 106 ms 104 ms 105 ms 0.so-7-0-0.XL2.DCA5.ALTER.NET [152.63.42.189]
10 104 ms 105 ms 105 ms 0.so-7-0-0.BR1.DCA5.ALTER.NET [152.63.43.177]
11 102 ms 104 ms 102 ms p16-0-1-3.r20.asbnva01.us.bb.gin.ntt.net [129.25
0.9.69]
12 100 ms 105 ms 101 ms ae-0.r21.asbnva01.us.bb.gin.ntt.net [129.250.2.1
7]
13 141 ms 140 ms 141 ms p64-1-1-0.r20.dllstx09.us.bb.gin.ntt.net [129.25
0.5.24]
14 145 ms 145 ms 147 ms p64-1-0-0.r20.hstntx01.us.bb.gin.ntt.net [129.25
0.4.71]
15 147 ms 145 ms 145 ms xe-4-1.r03.hstntx01.us.bb.gin.ntt.net [129.250.2
.229]
16 144 ms 144 ms 144 ms ge-9.ev1.hstntx01.us.bb.gin.ntt.net [129.250.11.
134]
17 144 ms 144 ms 144 ms gphou2-209-85-1-108.ev1.net [209.85.1.108]
18 143 ms 144 ms 144 ms ev1s-209-85-59-35.ev1servers.net [209.85.59.35]
Trace complete.
Ian J
I've used http://www.abeltronica.com/velocimetro/ ... newlang=uk earlier tonight to test speed several times over a 20 min period starting at around 9:30 pm.
I'm on a 4Meg cable link and got speeds varying from 1Mbps to 2.9Mbps. Not all that good on the face of it.
I've just repeated the test a single time and got 3.64Mbps....a value I've had consistently since being upgraded to 4Meg.
I'm on a 4Meg cable link and got speeds varying from 1Mbps to 2.9Mbps. Not all that good on the face of it.
I've just repeated the test a single time and got 3.64Mbps....a value I've had consistently since being upgraded to 4Meg.