[dyn!o]

Ok, I think I did my homework. I've tried to connect various Chineese places and monitor packets route/response/loss.

One of the servers I tried to connect is maintained by Shanghai Global Network. ExeTools is being hosted by China Mobile Communications so let's call them respectively SGN and CMC. As we already knew CMC seems to cause strange delays while connecting European users (I am assuming our case). When I connect to SGN then I (my packets) encompass almost whole Earth (X->Germany->USA->China - via the ocean) . When I connect to ExeTools then I am being routed on much shorter path (X->Germany->China).

Theoretically I should achieve much faster reply (not to mention bandwith) from ExeTools than the other server (hosted by SGN). And here the story begins. While connecting to ExeTools I lost the packets in six places. That's far too high. 1-3 would be ok but 6 problematic routers aren't something usual. Furthermore I have the following packets percentage losses:

1 - 20, 2 - 10, 3 - 10, 4 - 10, 5, - 20, 6 - 35.

So, the most interesting point is the sixth one. It misses so many packets (35%) that I was unable to quickly retrace correct ExeTools packets route because after this point I don't know the exact route but just see ExeTools response, without knowing what are the missing connections between sixth problematic router (211.141.237.27 - owned by CMC) and ExeTools itself (CMC too). Maybe there is a way to reveal the missing part of trace but I'm definitely not any network specialist.

Summarizing: 6 weak points and huge loss (35%) in one of them. Could it be responsible for ExeTools delay we encounter? Maybe, but if so then, theoretically, other (Chineese, US, etc.) ExeTools users should face similar problems (less sensible) since at this stage we are analyzing Chineese routers. So, going step further it would be nice to check each point (router) separately to reveal what is really going on. I used Ethereal - the very best sniffer available (well, I know, it doesn't support protocols rebuilding/recompilation but who cares in that case).

218.200.252.129 - China (CMC)
10/10, min = 2052ms, max = 2060ms, avg = 2057ms

62.159.199.166 - Germany (Deutsche Telekom)
no response, but I tricked it a little and the score was: 1417ms

Now that's something very wrong. 1417ms is a terrible response, not to mention that I'm located ~800km from this German router. Chineese 2057ms sounds like a joke too but let's assume it's ok (it isn't).

For those of you who don't want to waste the time on reading all the crap I wrote: I'm suggesting that these two places are responsible for connection problems with ExeTools. If your packets addressed for 211.141.237.27 (ExeTools) go by 62.159.199.166 then they will reach 218.200.252.129 too and that's the reason of our problems. If you're connecting from other continent than Europe you might not encounter the problems we do; the same if you're in Europe but your last national router doesn't reach 62.159.199.166.

Please notice I am not any specialist (nor even experienced) in networking topics so don't blame me for bad naming or wrong conclusions. It's only my small, humble hypothesis. Hope it was correct and helped at least one to understand ExeTools connection speed problem.

Regards.