Yes, hence I pointed out my VM connection's minimum is less than 0.1ms different to my BT one. Not sure what your point is? Most people's average is lower than my minimum too, then again most people don't live in Scotland.
---------- Post added at 00:26 ---------- Previous post was at 00:24 ----------
Quote:
Originally Posted by thenry
ive seen afew infinity connections having hjgjer min than VM even if its a little it is a little confusing with everything else looking good. i guess your cant win. not sure how mine will look.
|
Part of it I guess comes down to core network structure and routing. But as I say, mine's almost exactly the same on both, and I'm almost as far away as you can get and still make a comparison between the two. It'll also vary between sites. Right now for example, my BT Infinity minimum ping to Multiplay.co.uk is about 1.5ms slower than VM, but to BBC it's 0.5ms faster and to ThinkBroadband's ping monitor it's 4ms slower. Yet earlier in the day the trends were reversed. Still, the 4ms difference to pingbox1.thinkbroadband.com - i.e. the address used for the ping monitor(s) - seems to agree with your seeing BT plots being generally higher than VM's. Yet as demonstrated with the other two examples, similar differences don't occur to all destinations.
What really matters is average and standard deviation though. For that, the Linux built-in ping tool is far better than the Windows one. Here's some examples from the VM line:
--- www.bbc.net.uk ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 101700ms
rtt min/avg/max/mdev = 19.436/29.341/154.849/10.760 ms, pipe 2, ipg/ewma 101.802/30.844 ms
--- www.multiplay.co.uk ping statistics ---
1000 packets transmitted, 999 received, 0% packet loss, time 101483ms
rtt min/avg/max/mdev = 18.479/27.378/161.304/9.393 ms, pipe 2, ipg/ewma 101.585/26.287 ms
--- pingbox1.thinkbroadband.com ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 100951ms
rtt min/avg/max/mdev = 17.000/24.436/78.188/6.391 ms, ipg/ewma 101.052/27.891 ms
The 'mdev' value is quite good to get a measure of congestion to a particular host, as is the delta between minimum and average. Looking at the above numbers, the lowest congestion is seen along to pingbox1 - looks like VM has got a lot less congestion and a lot more unused capacity (or just a much bigger pipe) going to ThinkBroadband's network than, say, BBC's network. I wonder why... :-P
BT line, for reference:
--- www.bbc.net.uk ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 100922ms
rtt min/avg/max/mdev = 18.955/19.487/21.367/0.271 ms, ipg/ewma 101.023/19.532 ms
--- www.multiplay.co.uk ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 100898ms
rtt min/avg/max/mdev = 19.996/20.464/21.133/0.144 ms, ipg/ewma 100.999/20.457 ms
--- pingbox1.thinkbroadband.com ping statistics ---
1000 packets transmitted, 1000 received, 0% packet loss, time 101040ms
rtt min/avg/max/mdev = 21.242/21.737/22.869/0.183 ms, ipg/ewma 101.141/21.741 ms
Averages barely 0.5ms off minimums (vs. VM's 10ms, even off-peak), with standard deviations also well below half a millisecond. Sure, the distance adds to the minimum but at least there's nothing keeping me away from said minimum all the time