Page 1 of 1

Connection Times

Posted: Sun Oct 03, 2010 9:01 am
by dawmail333
Since I upgraded to ADSL2, I seem to be running into delayed connection times to all websites. Below I have a few listed ping times... are these normal values?

Code: Select all - Minimum = 40ms, Maximum = 42ms, Average = 40ms - Minimum = 35ms, Maximum = 38ms, Average = 36ms - Minimum = 295ms, Maximum = 296ms, Average = 295ms - Minimum = 210ms, Maximum = 215ms, Average = 212ms
Had really noticed that google pings were low before now, but compare the bottom two sites with the top two.
Is there some setting on my new router that might be causing this? I set up the adsl connection via trial and error, as the wizard was quite unhelpful. I'm running a D-Link DSL-2740B , firmware version 2.18.

If it would help, I could do some tracert commands, just ask.

Appreciating all help!

Re: Connection Times

Posted: Sun Oct 03, 2010 11:50 am
by CoreyPlover
Yes, those match my ping times almost exactly so they seem normal to me.

Google has some sort of peering with Verizon (I think) and Microsoft, and all Akamai content is cached in Australia too. Facebook,, etc each resolve to the west coast USA, hence the larger latency

Re: Connection Times

Posted: Sun Oct 03, 2010 12:46 pm
by dawmail333
Ok, thank you. Just seemed slower.
Spose it must be an illusion caused by having much faster DOWNLOAD times :D (just went from 1.5mbps to 13mbps!)

Thanks for that.

Re: Connection Times

Posted: Sun Oct 03, 2010 1:23 pm
by CoreyPlover
I don't think it is an illusion. It may have something to do with the exact routing at play, and also some component relating to the fact that ADSL1 = fast path and ADSL2+ = interleaved.

In ADSL2+, requested packets are "interleaved"; the packets are broken down and mixed together before being sent. What this does it make the transmission more error tolerant, which is usually necessary at such high speeds. But it means that time to completion of any packet is longer. See the Example and Disadvantages sections of (and also, ... 20#1215979 for an equivalent explanation)