Interpret the output of these traceroute commands

I've been getting unusually high ping (~180ms) to the League of Legends OCE server (Australia) when playing from NZ, and so I've done a traceroute for the IP of the server.

Is anyone able to give some insight as to what might be causing the problem? I've tried contacting my ISP but they have apparently "checked the line" and couldn't see a problem, and have otherwise been unhelpful. Also unusual is that this is happening specifically for the League of Legends OCE server (NA server has ~250ms, which is typical), as for other servers / games hosted in Australia I have ~30ms.

Below are the tracert outputs for both the OCE and NA servers.

OCE server tracert output:

OCE server tracert output

NA server tracert output:

NA server tracert output

2

1 Answer

Routes on a network can be asymmetric – the Internet is a mesh, and your requests may travel through one set of ISPs and carriers and return through another. And you can't actually see the return route using traceroute, unless you have some way to run it from the other side as well.

(Some companies have special "looking glass" webpages where you can enter your own IP address and have the ISP's system run a traceroute back to you – but Riot Games unfortunately does not.)

In addition, each one of the hops will make its own choice as to where to send the reply through: the Traceroute responses from Telstra back to your ISP take a shorter path, but the ones from VOCUS or Megaport take a longer one. Hence the sudden jumps up and down in latency along the way.

This might be because your ISP deliberately makes the normally-shortest path appear longer than others – maybe it wants to shift inbound traffic from an overloaded (or expensive) connection with carrier A to an underused connection with carrier B.

For example, when looking at the Megaport "Looking glass" webpage, the return path back to Compass can go either 1 hop through VOCUS, or 1 hop through Devoli, or 2 hops through Spark then Voyager. However, the two "short" routes are deliberately made to appear longer, so that the "long" Spark→Voyager→Compass route would be taken instead. And that might very well add up to 2x or 3x the normal latency.

(Note that I'm only making this guess based on pressing various buttons online and have no actual way of knowing why your packets really take this path or another – I'm not working for them. As such, this might very well be total garbage.)

1

Your Answer

Sign up or log in

Sign up using Google Sign up using Facebook Sign up using Email and Password

Post as a guest

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

You Might Also Like