hi again, the principle of routing is based on the *least cost of a route* where bandwidth, number of hops, latency, priorization are some factors. remember the internet has been designed as a failure tolerant mesh.
our servers in vienna have uplinks to 3 major backbone providers available, our server in US 2 - to some extent we can ask for priorizing one route or another but we don't have influence on peering points of the destination network.
eg. if DTAG decides to peer incoming traffic mainly with cogentco we actually don't get another possible route ...
anyway, thanks for reporting - we always keep a wary eye on such issues, christian
edit: hehe ... our provider managed to change the preferred route from our side and it should be much better now ...
and remember: a CRAY is the only computer that runs an endless loop in just four hours ...