Diagnosing Latency Issues

[dirtphillips]

When playing with friends in other places, the 3 people that play ddo in our house have identical latency issues around the same times (whether in same party or not). I already read the sticky related to dealing with lag issues and most of the fixes didn't seem related or I have already tried them.

I tend to think that it is either the wireless router or our cable internet connection (Hawaii). How do I determine the location of the game servers, as I would like to run a trace to determine the latency to the servers and see how many hops take place. What is this address? Also, does anyone else have any diagnostic tips on how to narrow down the cause? The only idea we have is to plug the computer next to the router directly into the modem when this next happens (tends to happen off and on about 50% of the time these days) and see the performance difference (if any).

Any help is appreciated, the 3 VIP accounts we have are going to waste and DDO is really unfun with latency above 1000 ms.

[Flavilandile]

If your connection to the mainland is through satellite then 1 second of propagation time ( there is no such thing as latency ) is quite good actually ( usually satellite bounces are around 1.6 seconds ).

Now traceroute ( tracert ) is the tool to use.
It will give you all the hops and such.

now the IPs to use are in the following lists :

74.201.102.43 ddo.com
74.201.102.43 www.ddo.com
74.201.102.13 gls.ddo.com
74.201.102.13 gls-auth.ddo.com
74.201.102.42 forums.ddo.com
74.201.102.11 patch.ddo.com

These seems to be the Server ones :

74.201.106.53 Wayfinder [DE]
74.201.106.64 Ghallanda
74.201.106.56 Argonnessen
74.201.106.21 Thelanis
74.201.106.22 Sarlona
74.201.106.23 Khyber
74.201.106.59 Cannith
74.201.106.60 Orien

First thing is to use a cable to connect to your router, as WIFI, like all the wireless protocols use Time Division between all the receivers. ( That is it split the time between all the flows it has to handle, so a 36Mbit/s Wireless connection split between 3 computers will be in truth a 12Mbit/sec connection for each computer. )
Wireline use Carrier Sense, if no carrier is seen on the wire a packet can be sent.

As for the Server locations they are in the Boston area IIRC. ( once upon a time I could be more precise... In Europe they were in AT&T Datacenter in Amsterdam. )

[dirtphillips]

Here is one example when I was having about 800 ms of latency, off and on for an hour. Server is Orien. Both of us online had similar issues, other players in group on mainland experienced no such problem.

1 1 ms 1 ms <1 ms 192.168.1.1
2 47 ms 65 ms 85 ms cpe-75-80-200-1.hawaii.res.rr.com [75.80.200.1]
3 144 ms 162 ms 128 ms tge17-3.milnhixd-cer03.hawaii.rr.com [24.25.233.113]
4 189 ms 188 ms 186 ms tge0-5-0-0.hnllhiik-ccr01.hawaii.rr.com [72.129.45.104]
5 183 ms 147 ms 143 ms agg21.milnhixd-ccr01.hawaii.rr.com [72.129.45.42]
6 260 ms 142 ms 64 ms agg31.lsancarc-ccr01.socal.rr.com [72.129.45.0]
7 65 ms 63 ms 71 ms ae-5-0.cr0.lax00.tbone.rr.com [66.109.6.102]
8 125 ms 85 ms 66 ms 66.109.9.24
9 121 ms 76 ms 71 ms te0-7-0-19.ccr21.lax05.atlas.cogentco.com [154.54.11.133]
10 94 ms 62 ms 63 ms te0-2-0-2.ccr21.lax01.atlas.cogentco.com [154.54.3.141]
11 903 ms 894 ms 876 ms te0-1-0-3.ccr21.iah01.atlas.cogentco.com [154.54.44.237]
12 1110 ms 984 ms 1048 ms te0-1-0-6.mpd21.atl01.atlas.cogentco.com [154.54.2.145]
13 129 ms 128 ms 159 ms te0-5-0-7.ccr21.dca01.atlas.cogentco.com [154.54.42.193]
14 140 ms 128 ms 129 ms te0-0-0-14.mpd21.jfk02.atlas.cogentco.com [154.54.42.34]
15 194 ms 178 ms 179 ms te0-1-0-2.ccr22.bos01.atlas.cogentco.com [154.54.44.38]
16 205 ms 181 ms 182 ms te3-2.ccr01.bos06.atlas.cogentco.com [154.54.46.130]
17 136 ms 162 ms 140 ms 38.104.252.70
18 171 ms 135 ms 174 ms border1.te8-1-bbnet2.bsn003.pnap.net [63.251.128.107]
19 154 ms 145 ms 194 ms turbine-7.border1.bsn003.pnap.net [64.95.76.202]
20 154 ms 143 ms 139 ms 74.201.102.154
21 176 ms 155 ms 279 ms 74.201.102.13

Trace complete.

[Flavilandile]

Here is one example when I was having about 800 ms of latency, off and on for an hour. Server is Orien. Both of us online had similar issues, other players in group on mainland experienced no such problem.

Well you have two obvious culprits :

11 903 ms 894 ms 876 ms te0-1-0-3.ccr21.iah01.atlas.cogentco.com [154.54.44.237]
12 1110 ms 984 ms 1048 ms te0-1-0-6.mpd21.atl01.atlas.cogentco.com [154.54.2.145]

I can make educated guesses about the router locations : lax sounds like LA, dca sounds like DC area, jfk sounds like New York, and bos is definitely Boston.
Well as now it is obvious that cogent uses IATA codes for their routers : iah is Houston and atl is Atlanta.
Now with that kind of problem, located where it seems to be located you have about 0 (zero) chance of getting it fixed. It is not your internet access, it is not your ISP, it is not the link between Hawaii and mainland and it is not a problem at Turbine side.

Looking at Cogent network map : http://www.cogentco.com/files/images...obal_large.png
my guesses seems about right for the router locations.

[MsEricka]

mmm Cogent.. the favorite carrier of all cheap ISP's and hosts. Very low cost data network, very high traffic.

Cogent is universally disliked by webhosts due to their latency issues. Unfortunately they are almost always used by Data Centers as a secondary or tertiary carrier due to their low cost and high line availability.

[Flavilandile]

mmm Cogent.. the favorite carrier of all cheap ISP's and hosts. Very low cost data network, very high traffic.

Cogent is universally disliked by webhosts due to their latency issues. Unfortunately they are almost always used by Data Centers as a secondary or tertiary carrier due to their low cost and high line availability.

Sadly it seems that all the DDO traffic go through them now.

I don't remember using them but a traceroute I just made shows that now I do.
I clearly remember having a Traceroute that went Paris -> Lanion -> New York -> Boston and using Equant network up to NY. ( I don't remember who it was on the NY->Boston Leg )

Now I have a trace that goes Paris -> London -> Boston through Cogent.

I suspect PNAP ( that seems to be the hosting company for Turbine servers ) changed their peering agreements, and now they only have cogent as peer.

[Missing_Minds]

I'll have to see how I'm getting through currently.

I know I've got other net issues that are not Turbine nor cog related. *sigh* someone is introducing signal loss on my node again. farking thieves.