Packetloss/lag unique to 1 game

hellzscream

Distinguished
Jun 12, 2007
139
0
18,680
For the past few months i've been having crazy packet loss with a game called Wildstar. It mostly occurs during prime time 7-10est. It is very sporadic it comes and goes my ping can be fine at 80ms for 10-30mins then climb up to 300-700ms for half an hour or so. It makes the game unplayable as skills don't register and my character isn't where it appears on my screen cause of the lag. I never get disconnected because of this issue. This is only an issue with this game not sure if it's a datacenter being overloaded during prime time? I've ran a tracert during the issue


Tracing route to 64.25.35-100.ncsoft.com [64.25.35.100]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms 192.168.0.1
2 * * * Request timed out.
3 11 ms 12 ms 11 ms 209.148.245.217
4 10 ms 11 ms 20 ms 69.63.248.181
5 * * 37 ms 64.71.241.110
6 24 ms 26 ms 27 ms xe-0-1-0.er2.iad10.us.above.net [64.124.196.213]
7 126 ms 125 ms 127 ms xe-3-3-0.GW12.IAD8.ALTER.NET [152.179.50.245]
8 * * * Request timed out.
9 162 ms 149 ms * GigabitEthernet6-0-0.GW12.DFW9.ALTER.NET [152.63.97.94]
10 158 ms 162 ms 160 ms 65.205.119.210
11 163 ms 164 ms 163 ms 64.25.35-100.ncsoft.com [64.25.35.100]

Trace complete.
 

hellzscream

Distinguished
Jun 12, 2007
139
0
18,680
I'm the only 1 on this connect directly connected to the gateway. Hitron CGN2-ROG is the model name.

result/3947774241]
3947774241.png
[/URL]
 
You main problem is your traceroute pretty much show no issues other than hop5 which may be nothing.

The ping times to the end devices are very good with almost no variation in this traceroute.

What you need to do is pick a couple of these making sure you have the first 2 and then what ever others you like. You then open multiple cmd windows and run continuous ping commands to each of these. The goal being when you see issues in your game you should be able to quickly swap to these cmd windows and stop them. You are going to hope that you see large increases in the ping times that correspond to what you games says.

If you are lucky...in a way..the problem will be in the first hop which means it is your router or something in your house or the second hop which represents the connection between your house and the ISP. Both these you have some control over getting issues resolved.

Now lets say the problem is in hop 7. This is the connection between abovenet(now called zayo) and alter.net (which was UUNET but is now verizon). What are going to do call the help line up for Zayo or verizon and tell them they have a network issue. You would not even get past the level 1 ticket taker because they first thing they are going to ask for is the circuit number which there is no way to find out.

So if you are lucky the problem will be in the first couple of hops. You only other solution will be to get a different ISP. Problem is the third hop is a rogers owned ip/router and if you are in canada rogers has control of a lot of the internet infrastructure so it may be hard to not use them.
 

hellzscream

Distinguished
Jun 12, 2007
139
0
18,680


I keep running tracerts during the time the issue arises and it constantly goes to high ping 150+ when it reaches iad8.alter.net. I'll try contacting verizion about this I guess
 

They will just laugh at you for even attempting to call them. You are not their customer you are not paying them anything so why would they even think to talk to you.

Be very careful the increase in latency itself means nothing as long as it is constant. If it is sometime 150 and other 500 then there can be a issue but when it is a fixed number it pretty much represents the distance your data must travel. Stuff does not always take the most direct route and even worse the path from you to the server may actually be different than the path from the server to you. Traceroute only shows the path from you to the server.



 

hellzscream

Distinguished
Jun 12, 2007
139
0
18,680
i've been getting better and more stable ping lately did a tracert and it seems i'm going through level 3 servers now. Is the path my connection takes just random? Or is it based on traffic which then gets throttled?

xH4tMWB.png
 
It is not random but the ISP do change things from time to time to improve the performance of their network. There also could have been some connection down that was repaired. And there is some random to it because if the ISP are not very careful their routers will pick the first path to a destination it learns and as time goes on and you get small circuit drops this will change. This is poor BGP design but some ISP are not always the smartest.
 

hellzscream

Distinguished
Jun 12, 2007
139
0
18,680
ran test using pingplotter huge packet loss

Target Name: 64.25.34-201.ncsoft.com
IP: 64.25.34.201
Date/Time: 17/12/2014 7:06:57 PM to 17/12/2014 7:13:12 PM

Hop Sent Err PL% Min Max Avg Host Name / [IP]
1 26 0 0.0 0 6 0 [192.168.0.1]
2 26 0 0.0 7 17 9 [7.11.162.237]
3 26 0 0.0 8 31 12 [209.148.245.221]
4 26 0 0.0 8 25 13 [69.63.248.173]
5 26 0 0.0 9 26 13 [69.63.248.181]
6 26 7 26.9 23 26 24 [69.63.249.26]
7 26 0 0.0 23 73 29 xe-0-1-0.er2.iad10.us.above.net [64.124.196.213]
8 26 1 3.8 122 146 127 xe-3-3-0.GW12.IAD8.ALTER.NET [152.179.50.245]
9 26 26 100.0 0 0 0 [-]
10 26 1 3.8 149 178 154 GigabitEthernet7-0-0.GW12.DFW9.ALTER.NET [152.63.97.98]
11 26 0 0.0 156 188 161 [65.205.119.210]
12 26 0 0.0 147 186 153 64.25.32-9.ncsoft.com [64.25.32.9]
13 26 2 7.7 149 163 151 64.25.32-26.ncsoft.com [64.25.32.26]
14 26 0 0.0 149 172 152 64.25.34-201.ncsoft.com [64.25.34.201]

j2y8Ll4.png
 
The node that show most the loss is in rogers network but that is not always a issue if the router itself was having a issue but was passing traffic ok. If for example hop 9 was actually losing 100% no traffic would ever hit the hops after that you would have 100% from that point on. So if hop 6 was actually dropping 26% of the packets then all hops past there should also lose 26% or more.

A much larger issue is the increase in latency between hops 7 and 8 which is between abovenet and verizon. That implies the pathing of the traffic is not optimum. You would have to get the cites to see but it could be going all the way across the country.

What is really bad about trace router is it only show you the path from you to the server not the path back. It can come back on a completely different path which makes troubleshooting even harder when you have no access to equipment on the other end. It is very common for it to run asynchronously because a ISP will always prefer to use their own network rather than a competitors. So each end can have a very different path in some cases especially when more than 2 ISP are involved.