Guild Wars 2

SERVER : BLACKGATE

PROBLEM : LAG LAG LAG , kittening skill lag, on SMC 3 way fight , and every single player complain about this ! we payed for this game , DO SOMETHING !

kitten I N G FIX THIS YOUR kittenS !

for kittening 4 years straight up the same kitten problem come on !

if you guys don’t know , how to solve this ! please hire www.level3.com

#BRAZIL #HUEHUEHUHEU #SOUTHAMERICA #TOCANTINS #INDIO’SDAY #TILAPIA
Ty!

It was us on CD that caused the lag!!!!! WE REFUSE TO USE DVDS…MUAHAHAHAHHA.

Anet need to add more cleave+AOE spam class capabilities so that becomes “normal” gameplay.

just wait until next expansion…

It’s important to make the distinction of what “lag” means. I know you’re referring to skill lag…

ANet support will have you use WinMTR. You might see packet loss at an intermediary node between you and NCSoft’s services (in my case, the Comcast → Level 3 handoff). This is not properly defined packet loss. It is the control plane of the network device dropping packets. This can be made clear by using ping (ICMP) to that specific device that shows packet “loss” to the specific intermediary node.

A great in-depth presentation on traceroute can be found at https://www.nanog.org/sites/default/files/tuesday_steenbergen_troublshootingtraceroute_62.49.pdf.

Take a look at page 33, especially the last bullet point. And again on 36, pay special attention to the first and third bullet points.

So while your WinMTR trace might look something like:

MyISP – 0% packet loss
Level3 – 50% packet loss
NCSoft_service – 0% packet loss

What ANet support, and by extension NCSoft network engineers, will tell you is that you’re experiencing packet loss and there is nothing they can do about it as it is “your” issue. Of course, this doesn’t indicate packet loss what so ever. We’re not dropping packets at NCSoft’s services. We’re dropping them, in the above example, at Level3 because traceroute packets are just not a priority for those network devices — nor should they be.

If we were having true lag, say for example you had a poor network connection and were overseas, you would see “rubberbanding”, no skills executing, possible disconnects, and so forth.

And another indication that this is not a packet loss issue is that Skill #1 never fails to go off. Skill #6 (heal) will most often go off. It is all of those other skills that are a toss-up of whether they’ll go off with a delay or simply outright cancel execution. You won’t see rubberbanding, character movement is fluid (sans FPS drop), Skills 1 and 6 work…

This is indicative of a server-side prioritization and potentially performance issue. Of course saying “performance” is too generic as we don’t know what the back end looks like. They could possibly have the best performing servers money can buy, but if the GW2 server software isn’t designed for these massive 2 to 3-way engagements, it’s not going to perform well regardless.

In the end, there’s nothing you nor I can do. It’s not a network issue, despite what NCSoft network engineers indicate. But as they’ve been insistent it is a network issue, there’s nothing to resolve. It’s simply a lack of understanding of how traceroute functions and not understanding how traceroute works.

Ni In.6578:

This is indicative of a server-side prioritization and potentially performance issue.

Uhm… we already know what it is because the devs told us. In simple terms its the server skill queue that chokes on requests and is unable to respond to skill activations. The skill queue is separate from the movement, thats why you dont necessarily rubberband when skill lagging (or vice versa).