MilwaukeePC
-
I generated this data using these settings in WireShark's IO graph:
I wasn't doing anything special during this test. What can I do to figure out what's going wrong when I try to connect to the EU servers?
The IPs that exhibit this seem to have RDNS matching
ec2-18-*.eu-central-1.compute.amazonaws.com
-
@ben_lubar said in MilwaukeePC:
What can I do to figure out what's going wrong when I try to connect to the EU servers?
Hack into your nearest trans-oceanic fiber pipeline and sniff the shit out of the pungent miasma that emanates therefrom?
-
Well, I sent the data to the other end of the connection, so maybe they can help?
-
@ben_lubar said in MilwaukeePC:
What can I do to figure out what's going wrong when I try to connect to the EU servers?
Can you use a traceroute to figure out where the packets are going AWOL?
Of course, since it's going into AWS then the problem might be not anything anyone running the sites can comprehend because it might just be congestion on the network to other co-hosted VMs. (Also, you could ask some of us on this side of the pond to replicate the network tests and see if we can at least discount issues on the US side or the transoceanic link; reaching one the US AWS instances is unfortunately not enough for that to be certain because backbone routing is complicated.)
-
Doesn't the title of this topic answer the question?
-
I presume you've tried this, but the windows utility pathping pings the entire route, and the loss per node