OldCoaly wrote on Jan 4
th, 2015 at 7:07am:
The different game worlds have different lag profiles.
Next time you're rubberbanding, log into another server with a different account and compare, keeping the original connection open.
Some lag comes from pinch points on the route and impacts all servers, but the debilitating stuff that people are complaining about right now is not happening to all the servers.
I've seen weirder things caused by a hardware issue on a router...
Things like : One single IP address on a customer network being unreachable if you came through a given router on the ISP backbone... if you came through any other router in the backbone the issue was not there... in the end it was caused by an issue on the card facing the customer network. ( all the other interfaces on that card worked fine, it was impacting one single interface out of 10. ) All the other addresses on the customer network were reachable from everywhere. ( the customer has a /28 on that network one single address couldn't be reached, all the other addresses were... )
And it depends on how the actual game server ips are advertized in the network ( read : routed ).
If they are advertized as unique /32 then it's quite possible to have a huge issue on one, and no issue on the next address/server.
( or use my above example and instead of a 100% loss for a given IP take a 5 to 50% variable loss, you'll have the Lag effect )
I can't legally check how they are advertized right now... ( I have no reason to connect to my customer nodes on a sunday afternoon while still being in vacation )
but I might be able to steal a few minutes to perform some checks tomorrow once I'm at work.