PDA

View Full Version : Bandwidth Meters



nigelbleddfa
6th December 2009, 07:17 PM
I am currently conducting some tests with a bandwidth meter. The one I am using is Shaplus 1.3. which is available as freeware. I am using it to compare bandwidth usage of four different trading softwares and the results obtained so far are very interesting.
Does anybody know if the meter I am using is anywhere near accurate ? Has anybody else considered performing similar tests, please ?

Anonymous
6th December 2009, 07:32 PM
Have a read of Cran's blog tests:

http://cranthetrader.blogspot.com/2009/10/bandwidth.html

http://cranthetrader.blogspot.com/2009/10/bandwidth-2.html

http://cranthetrader.blogspot.com/2009/10/compare-meerkat.html

nigelbleddfa
6th December 2009, 08:02 PM
Thank you. My meter shows that The Toy is using somewhere in the region of 75 MB per hour. I am unable to reconcile my figure with that shown at the sites you have provided. Is my figure accurate, please ? The meter seems to be simple to use but I could still be doing something wrong.
I will run all four of my softwares separately for one hour tomorrow morning when I will be alone at home so ensuring that the only softwares using my broadband connection are the trading ones. They will all be set to the same market which will be in-play much later in the day.
Regardless of whether my figures are correct in absolute terms, the test should give an idea of the relative use of bandwidth by each.
I would be grateful to be corrected if this method is likely to be unreliable.

The Geek
6th December 2009, 08:05 PM
The results obtained so far are very interesting.

It will be interesting to see your results.

The Geek
6th December 2009, 08:13 PM
I believe Crans monitor is in kbps. { Kilo bits per second. } I think he put Mb/s in error looking at the pics.

Anyway to get to MBytes per hour form kbits ps....

100kbps * 60 seconds * 60 minutes = 360000 kbits PH

360000 / 8 = 45000 kilo bytes PH

45000 / 1024 = 43.95 Megbytes Per Hour.

So your figure of 75MB would be correct. Really depends how you have TT configuerd. I checked the current version of Nicko a little while back, and a default install was using over double the bandwidth of TT.

nigelbleddfa
6th December 2009, 08:24 PM
You know infinitely more about this sort of stuff than I do ! Is my test likely to give a proper indication of relative bandwidth usage, please ? Is my thinking essentially correct or is it fundamentally flawed ?
I have already compared two softwares but these tests were conducted when my wife and daughter were using Internet connections at the same time. My initial observation is that one software uses twice as much bandwidth as one other. The other two are not yet monitored.
One question immediately springs to mind. Are you already aware of the results I am likely to find ? I am "scientifically trained" as a chemist not a programmer.
I'll do my best and provide honest results. Nobody should rely on my findings and I would advise others to satisfy themselves as to the true figures.

The Geek
6th December 2009, 08:35 PM
I have only tested Nicko vs TT. Cran did a few.

I guess you are likely to find the same as Cran & I.

TT by default installs on a Hybrid mode simialr to that used by most.

Evo uses complete, which sucks up a lot more bandwidth. You can set TT to complete if you so choose, but it is a complete waste of bandwidth for most types of trading.

nigelbleddfa
7th December 2009, 02:45 PM
I have read elsewhere that the Betfair Software has a small tick box which enables the user to cut back on bandwidth usage. I conducted my tests and was surprised to see that this particular software used much less bandwidth than the other three. Doubting the accuracy of my testing, I submitted a bet which was matched quickly, so the software appeared to be working properly. I carried out a second test and found that the second reading was slightly lower than my first. Clearly, it is different to the others in the way it works either in its own right or, specifically, on my machine.
The tests were carried out in a situation where my broadband was being used solely by the trading application. The same market was chosen for all, the Over 2.5 goals in tonight's Watford V QPR game. The grid function was used in all of the tests.
The bandwidth usages were as follows :

Betfair Software 15.4Mb/hr, 13 Mb/hr

Geek's Toy 45 Mb/hr

Bet Devil 65.5 Mb/hr

Betraider Nicko 124 Mb/hr

By observation I noted that the bandwidth usage of all four was constant with no signicant peaks and troughs.
I used Evolution on its first day of release. It was consuming a very large amount of bandwidth, over 2.5 Gb per day. The software was suspended as Betfair could not cope with its demands. It would seem that it uses much less bandwidth now than when it was first released.
My own ISP will be increasing my bandwidth allowance significantly at the end of January so bandwidth usage will not now be a determining factor in deciding which softwares I will continue to use. For others whose ISP's have a fair usage policy, there could be problems.

The Geek
8th December 2009, 10:16 AM
The only one that contradicts my curent understanding of bandwith usage is the first one. What Refresh speed did you have set on it?

The interesting thing about your test, is that on a 2 runner market Nicko would be using 124 Mb/hr.

The number of runners in the market significantly affectes bandwidth usage, so it would be interesting to see a comparison of the 4 products on say a 10 runner horse race, 5 minutes before the off, as that is when a lot of traders will be using it.

nigelbleddfa
8th December 2009, 12:58 PM
The only one that contradicts my curent understanding of bandwith usage is the first one. What Refresh speed did you have set on it?

The interesting thing about your test, is that on a 2 runner market Nicko would be using 124 Mb/hr.

The number of runners in the market significantly affectes bandwidth usage, so it would be interesting to see a comparison of the 4 products on say a 10 runner horse race, 5 minutes before the off, as that is when a lot of traders will be using it.


I wanted to compare the softwares using the same market. This would not have been possible if I had chosen a horse race. To be fair to each I picked a market which was not being traded heavily at the time of testing.
The Betfair software shows a refresh rate of 1.0. If I set to 0.5 it is possible it will change and I'll retry with this setting later. Nevertheless, I wouldn't expect the figure to change so as to bring it anywhere near the other three.
Previous usage of bandwidth by Nickolution has been been underestimated, but when it was first released it was using very large amounts of bandwidth, over 2.5 Gb per day ( These figures were provided by my very reliable ISP ) so it seems that there has been a significant mprovement.
I'll test the new version tomorrow, assuming I am able to download it at some time today.

The Geek
8th December 2009, 01:00 PM
Ah OK that explains it. I think the refresh rates on TT & BetDevil are 0.2, so that's why there was the big difference.