Woodlane

Please consider automatically logging a value of -1 for Upload, for example, when the user does not select Upload testing. Have average speed calculations exclude this 'place holder' value. Having some kind of 'place holder' entry of the unselected test will make Testmy Upload speed and Download speed points align above and below each other on our personal results graphs. Auto Test - log 'place holder' -1 when for unselected test Typo - Topic should read: " Auto Test - log 'place holder' for unselected test."

Does anyone know of a site or utility that can calculate a cpk or similar metric from download and upload speed measurements? A lot of us us are concerned about the variability of our speeds, but I have not seen a tool that will actually quantify it. Maybe Testmy.net could add this as a feature from data that is being collected combined with the speed that the member is paying for. In the meantine I am looking.

"as for IE, it seems to be sensitive and I often hear complaints about it being broken in some way or another."...I am not surprised since this is a Microsoft application. I usually use Firefox 13.0.1. I find that IE and sometimes Firefox act up especially when there are a lot of TCP Delays caused by 'traffic shaping'. The 600mS+ latency of satellite might also be a factor for me. When I get the time I will be moving to a Linux setup for monitoring. My hope is that sometime in the deep future rural high speed access reliability will be prdictable like picking up the phone or accessing my satelite TV programming. "I have just been curious not critical since I first started using TestMy, as to why IE8 is showing me speed results that are unusually higher than I have ever seen in 8 years," Possibly 'traffic shaping". I find that the ISP is able to adjust speeds dependent on the application. Typically time-dependent apps like VoIP are less affected than file transfer aps like FTP and BitTorrent. This is why I use these speed tests as tools to look at connection speeds when I experience sluggish performance. Sometimes there is a lot of disk access going on slowing my notebook unrelated to my broadband connection.

Working OK now. All I did was reset my notebook again. My connection has been a bit sluggish bringing up the results page lately, so maybe my connection could have been causing the problem....

For last few days I have not been getting a graph on the results page even after waiting a while. I have cleared cache, restarted the notebook and tried a smaller sample size..

This was happening for both auto and non-auto tests. My results started mysteriously being logged again this evening. ....case closed. Might have been caused by some kind of corruption of data that I was sending.

For some reason, since Tuesday night, my tests are not showing up in my history (username: 'Woodlane') even though I am logged in an recording as usual. I have restarted my notebook and tried again, but issue persists.

Mudman4..."...more precise.."..another option to a log scale might be the dual 'Y" axis scale to which you refer that would show the much lower value Upload speeds in more detail. Tdawnaz ...... a log is a scale type rather than a grap type. Bar and lines are graphing types. BTW... a possible bug in the existing graphing program.......the right end of download speed line is always shifted to the left of the upload spped line. The right ends of both lines are not under each other at the same data point.

Upload speed is often much less than download speed to the point where it appears as a straight line near the bottom of the graph. A log scale would show fluctuations in upload speed much better.

It would be really helpful to have these two percentages calculated and added to the report and exportable to csv file for the Automated Speed Test. This would record how consistent download and upload speeds are being delivered. I experience some TCP pauses which affect performance and this would be a way to track ISP's improvements to service.

I was hoping to save that data because it might be a valid burst on a small filesize reflecting potential download speed capability of this new satellite. My other option would be to save the csv, then delete the data point as you suggest. Maybe in a future release this scaling feature might be a possibilty.

Would be nice to have a user-settable scale so that the effect of 'flyers' can be ignored. Most of my readings are around 3mbps. A 57 mbps resuly has made it hard for me see changes in normal results.

Especially for satellite connections with 400ms + latency.

For myself, my ISP if he wants to understand the data better to implement better service, and for any others who want to analyse the causes of performance changes. All the more reason to include a QOS percentage calculation in the exported csv file. The csv is a valuable tool in the challege to improve broadband for everyone.

For the speeds that I work with, I find Testmy results compare favorably with tools used by a network engineer. I really appreciate Testmy's tools particularly the data logging and flexibility of selecting test file sizes. It works for me. I have submitted a suggestion to add a QOS calculation to the data log csv file. For me, that would be of great benefit in understanding variations in speed at different times of the day.

I would expect 'Next 25" to scoll the chart to the right displaying more recent data points if they exist, and "Previous 25" to scroll the chart to the left displaying older date stamped data points since these selection boxes are located below the X (Date and Time) Axis.. I currently have to select 'Next 25" to go backward in the data set. Could you consider a reversal in functionality of these selections in a future release?

I'm averaging 3mbps down, 400kbps up, 600ms latency on their new Viasat-1 service. Their old satellite was slow and cost much more $$$$. I'm on their standard package advertised at 3mbps / 500 kbps. I find their phone and email support very good. Doing Skype OK. No tower...phone freed up...no other options here...looking good. I'm just monitoring their connection reliability now usingTestmy.net. I'm averaging 3mbps down, 400kbps up, 600ms latency on their new Viasat-1 service. Their old satellite was slow and cost much more $$$$. I'm on their standard package advertised at 3mbps / 500 kbps. I find their phone and email support very good. Doing Skype OK. No tower...phone freed up...no other options here...looking good. I'm just monitoring their connection reliability now usingTestmy.net.

I'd be using the comment field to add occasional brief freeform comments related to the test like "Increased filesize to 2.5mb" or "Weather related degredation" or "Spot beam issue.".

Speed's great, but without a big bucket, it loses some of its benefit. Maybe there could be a similar poll for MB/month allowance or $$/MB.

Increasing the mininum upload test file size to 1.5 mb or greater has resulted in much more accurate upload speed results for satellite connections like this because of their high latency in the area of 600ms +.

Upload speed is now much closer to what the ISP is supposed to be providing. I have set the minimum upload file size to 1.5mb. That seems to work OK. It produces results over 400kbps up vs 200 before. 600ms+ latency was causing misleading results with small test file sizes.

Can the 'Testmy" Automatic Combined upload and download test have a selectable file size like 2.5mb so that it produces more accurate results for satellite connections? I tried a 2.5mb upload test just now and the result was much closer to what I am supposed to be getting with my service. The 128kb file size produces way low results for upload testing.

I am a rural resident and satellite is my only option here on an island in the 1000 Islands area of Eastern Ontario. My neighbors has had slow performance with the old $$$ xPlornet satellite service and a competitors wifi. Late last year xPlornet parntered with the new Viasat-1 4g satellite. Having an IT background, I felt that satellite was worth taking a look at since I am not a gamer, and the latency should not be an issue for me. Also cost has come down. This is my first month trying it out. It is OK for Skyping. I'm averaging 3mbps down but having probems getting consistent results across different testing sites. This site records around 220kbs up, a flash site records around 400 kbps up. I'm supposed to be getting 500 kbps up. Reliability could be better, I have had a couple of 24hr outages already for technical reasons and speed fluctuates, but I am trying to live with these for now with the hope that the bugs will be ironed out and 4g technology will come through. I'd like to see a way to add time and date stamping to the results graph here at Testmy, otherwise thsi is a fantastic site. Found it using Google. I find xPlornet customer service very good. The pre-sales supplier could have been a bit better reviewing the contract and billing process with me but their installers were very competent and got me a good SNR...averages around 8.5 sometimes hits 10.