How much data does the Speedtest integration download and upload for each running instance?
The release notes are not specific on this point. If I run speedtest (ookla-speedtest-bin) on the command line, eg. speedtest -s 12491 , which is to Telstra Melbourne,
I get :
Download: 105.61 Mbps (data used: 87.2 MB)
Upload: 18.96 Mbps (data used: 10.0 MB)
Is the integration the same ?
The speedtest integration is the same as the command line, but each instance might be different.
An ethernet connection do not know how much it can handle, so it starts in the low and then increase with each successful delivery. At one point some of the lines between your network card and the server will shot over the max capacity and then it will fail. Once it fails it will jump back quite a bit and start over with increases, but in smaller jumps than before. This way it will slowly inch closer and closer to the max. capacity.
The max capacity is not just your current connection, but all connection on the different lines between your network card and the server. It can be other connections to/from same netword card, it can be the movie streaming in the living room from the NAS in the garage, it can be the movie streaming from Netflix over the router and it can be the rush to see some national event on internet streaming from other users at your ISP, just to mention a few cases.
Each of these and many other usages will affect the length of the test, because if the inching closer to the max. capacity is suddenly interrupted by a huge Windows download, then Speedtest will get kicked back on the max. it had already reached, and it will be uncertain on the result, so more testing will be down and more data will be moved.
thanks for that detailed explanation.
Since as you explain, the integration is essentially the same as the command line application , it looks like I can expect a total data transfer of around 100mB per run, more or less.
That is good enough for my purposes. I’ll probably just script it to run avery few hours, on the odd minute.