Jump to content


Lag Problems (Detailed)


  • Please log in to reply
13 replies to this topic

Userext #1 Posted 16 October 2017 - 06:56 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

Hi there, I have been lead to forums by the customer service and was informed that they can't do anything about the lag problem as they can't send our ideas(?) to developers...

I don't know why they thought the lag problem caused by corebackbone (WG's internet service provider) and by servers directly, is somehow an idea that developers can do something about...

 

Anyway cut short but first some knowledge for those who don't understand these graphs so that I won't see anyone say "hey! that ip causes PL! ask them why it happens!"

 

Firstly, in any pingplotter graph you will most likely come across a hop with high ping and possible PL. This is most likely "gateway" and to confirm it, you'll have to look at hops before and after it. For example if hop 5 is a gateway there won't be much of a difference for ms between hop 4 and hop 6.

 

Secondly, in any pingplotter graph, if you spot a problem(PL, high ping etc) but it doesn't contunie up to final destination, then according to pingplotter it's a distraction by the demons of bad networking mamba jamba

Example:

 

-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

 

Now that you have learnt some information about how pingplotter operates,

This is my problem:

 

As you can see, my ping goes from 50 to 84 the second it reaches corebackbone. Now keep in mind that hop 7 is outside of my ISP's authority range and in europe that belongs to another company.

But that's not all! From 84 to 95 by the time it travels on corebackbone fibers and it reaches to 116 from 95 at server which has a 4.5% PL that is directly caused by server itself(no other PL all the way to server)

 

Here is the funny part, the picture below is my pingplotter result from today

Please notice the huge difference between charts from the point my connection reaches to corebackbone and then to the server. 66 to 69 and 84 to 116? This definetly shows two seperate problems.

 

-Corebackbone has an issue with their infrastructure.

-EU1 has a PL problem which I'm guessing caused by the low capacity of servers.

 

So far, customer service infromed me that people from my region and even from egypt complained yet not a single romanian complained.

Aside from the obvious fact that cs doesn't understand network coding, internet problems show theirselves at the far edges of connections more often and stronger than those who are closer to it. So if you don't have any problems with servers, that doesn't mean there isn't a problem. It rather means that a there are symptoms of a problem yet it hasn't effected the server fully yet.

 

 

Here's an additional information:

- Hop 3 is an internet speed control hop while bafra and bartin are both gateway hops.

- At the start of every month, internet speed controls get reset which means hop 3 can't cause problems to me at next month.

- At the start of this month until 1-2 days ago, I had an average of 45ms ping with 30 to 55 range.(caused by speed control)

 

 

 

Why am I posting this? Well these are symptoms of a problematic infrastructure by WG's ISP and/or problematic server(s). If I can provide more evidence and prove that servers are having problems, it'd be nice to have your support since the servers will degrade overtime and afterawhile you'll feel the effects aswell.

 

I can always fix the issues on my ISP's end by annoying them for weeks while WG's customer service just tosses me aside without reading the company's manual they are affiliated with.

 

 

I'll collect more data by checking my ping at the least crowded hours and at the most crowded hours for a week and add anything unusual(like ISP upgrades or repairs) to my report and will come back next week.

Please note that I didn't come here before Customer Service. I came here after they sent me to here.

 

Sources:

https://www.pingplotter.com/manual

https://www.pingplot...acket-loss.html


Edited by Userext, 16 October 2017 - 07:02 PM.


Red_Dragon_Firkraag #2 Posted 16 October 2017 - 07:20 PM

    Lieutenant

  • Player
  • 15476 battles
  • 1,682
  • Member since:
    08-13-2011

please provide a pingplotter screenshot of tests to the following servers:

92.223.1.102

92.223.1.123

92.223.20.214

 


Edited by Red_Dragon_Firkraag, 16 October 2017 - 07:21 PM.


iztok #3 Posted 16 October 2017 - 07:23 PM

    Major

  • Beta Tester
  • 35215 battles
  • 2,520
  • Member since:
    10-28-2010

Hi!

Do you realize WG has NOTHING with these bad-behaving 'net servers?

The only thing you can do is find the admin(s) of these servers and send them the pingplotter picture. And wait...

 

BR,  Iztok



Userext #4 Posted 16 October 2017 - 07:36 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

View PostRed_Dragon_Firkraag, on 16 October 2017 - 07:20 PM, said:

please provide a pingplotter screenshot of tests to the following servers:

92.223.1.102

92.223.1.123

92.223.20.214

 

 

How many counts and intervals?

edit:I'm going to go ahead and apply 100 count 2.5 interval.

 

View Postiztok, on 16 October 2017 - 07:23 PM, said:

Hi!

Do you realize WG has NOTHING with these bad-behaving 'net servers?

The only thing you can do is find the admin(s) of these servers and send them the pingplotter picture. And wait...

 

BR,  Iztok

 

Nothing? really? How many times more do you need me to repeat that corebackbone is WG's internet service provider? They have more power over their ISP than I do over mine.

 


Edited by Userext, 16 October 2017 - 07:37 PM.


Baldrickk #5 Posted 16 October 2017 - 07:40 PM

    Field Marshal

  • Player
  • 30507 battles
  • 14,623
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

Erm. I can't see your pingplotter to WG...

 

Am I missing something?


Edited by Baldrickk, 16 October 2017 - 07:41 PM.


Userext #6 Posted 16 October 2017 - 07:54 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

View PostBaldrickk, on 16 October 2017 - 07:40 PM, said:

Erm. I can't see your pingplotter to WG...

 

Am I missing something?

 

don't know why you can't see it. Does anyone else have the same problem?

Just in case

First pic

Spoiler

 

And second pic

Spoiler

 

View PostRed_Dragon_Firkraag, on 16 October 2017 - 07:20 PM, said:

please provide a pingplotter screenshot of tests to the following servers:

92.223.1.102

92.223.1.123

92.223.20.214

 

1.102

Spoiler

 

1.123

Spoiler

 

20.214

Spoiler

 



Baldrickk #7 Posted 16 October 2017 - 07:57 PM

    Field Marshal

  • Player
  • 30507 battles
  • 14,623
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostUserext, on 16 October 2017 - 07:54 PM, said:

 

don't know why you can't see it. Does anyone else have the same problem?

Just in case

First pic

Spoiler

 

And second pic

Spoiler

 

 

1.102

Spoiler

 

1.123

Spoiler

 

20.214

Spoiler

 

I can see those, thanks.

 

Yep.  There is a problem there.  

Unfortunately, you're unlikely to find an answer here, no idea why support tell people to come here. 


Edited by Baldrickk, 16 October 2017 - 08:03 PM.


Red_Dragon_Firkraag #8 Posted 16 October 2017 - 07:58 PM

    Lieutenant

  • Player
  • 15476 battles
  • 1,682
  • Member since:
    08-13-2011

Your ISP's (Turk Telekom) backbone is oversaturated. The situation is especially worse between hops 5 and 6. It's not WG's fault here.

 

Because the issue is inside your ISP's infrastructure, not even a gamer VPN service such as WTFast or Pingzapper would fix this.


Edited by Red_Dragon_Firkraag, 16 October 2017 - 08:03 PM.


Userext #9 Posted 16 October 2017 - 08:01 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

View PostRed_Dragon_Firkraag, on 16 October 2017 - 07:58 PM, said:

Your ISP's (Turk Telecom) backbone is oversaturated. The situation is especially worse between hops 5 and 6. It's not WG's fault here.

 

Because the issue is on your ISP's side, not even a gamer VPN service such as WTFast or Pingzapper would fix this.

 

But that's not what pingplotter says in their manual.

 

The final destination is most important

 

If the final hop is showing 0% packet loss and acceptable latency, you can ignore whatever you see in the hops leading up to it. In example 1, we see scary looking packet loss early in the route, but performance looks good at the final destination. We can ignore patterns like this.

 

https://www.pingplot...acket-loss.html

 

Edit: 1.123 is what I target here with this arguement.


Edited by Userext, 16 October 2017 - 08:05 PM.


Red_Dragon_Firkraag #10 Posted 16 October 2017 - 08:05 PM

    Lieutenant

  • Player
  • 15476 battles
  • 1,682
  • Member since:
    08-13-2011

View PostUserext, on 16 October 2017 - 08:01 PM, said:

 

But that's not what pingplotter says in their manual.

 

The final destination is most important

 

If the final hop is showing 0% packet loss and acceptable latency, you can ignore whatever you see in the hops leading up to it. In example 1, we see scary looking packet loss early in the route, but performance looks good at the final destination. We can ignore patterns like this.

Not correct. Some ISPs may falsify the result by prioritizing outside ICMP packets before packets destinied for the infrastructure itself. But here the issue is so clear it cannot be 'corrected' by falsification.



Baldrickk #11 Posted 16 October 2017 - 08:13 PM

    Field Marshal

  • Player
  • 30507 battles
  • 14,623
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostUserext, on 16 October 2017 - 08:01 PM, said:

 

But that's not what pingplotter says in their manual.

 

The final destination is most important

 

If the final hop is showing 0% packet loss and acceptable latency, you can ignore whatever you see in the hops leading up to it. In example 1, we see scary looking packet loss early in the route, but performance looks good at the final destination. We can ignore patterns like this.

 

https://www.pingplot...acket-loss.html

 

Edit: 1.123 is what I target here with this arguement.

I can't verify with a ping plot of my own (not at home) but it is pretty indicative that it's the hop where it leaves you ISP's network that you have the problem. 

 

You are correct about packet loss, unless it continues up the chain, it's not a problem.

Same with ping spikes. 

 

What we have here though is slow traffic into and out of the ISP's network on the internet side.

This is consistent and goes up the chain, hence the increase in the average ping.

 

The only thing you can do is talk to your ISP. 



Userext #12 Posted 16 October 2017 - 08:16 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

View PostRed_Dragon_Firkraag, on 16 October 2017 - 08:05 PM, said:

Not correct. Some ISPs may falsify the result by prioritizing outside ICMP packets before packets destinied for the infrastructure itself. But here the issue is so clear it cannot be 'corrected' by falsification.

 

I'm going to need you to find where it says that in pingplotter manual. There are mentions of ICMP but pingplotter only mentions low priority problems and that it shouldn't matter.

Example picture

Spoiler

 

If it's my isp's problem I can get it fixed.

 

Also I'm getting new/different readings on pingplotter now. I'm going to test all 3 of those pings again and post results of 100 count 2.5 interval



Baldrickk #13 Posted 16 October 2017 - 08:21 PM

    Field Marshal

  • Player
  • 30507 battles
  • 14,623
  • [-TAH-] -TAH-
  • Member since:
    03-03-2013

View PostUserext, on 16 October 2017 - 08:16 PM, said:

 

I'm going to need you to find where it says that in pingplotter manual. There are mentions of ICMP but pingplotter only mentions low priority problems and that it shouldn't matter.

Example picture

Spoiler

 

If it's my isp's problem I can get it fixed.

 

Also I'm getting new/different readings on pingplotter now. I'm going to test all 3 of those pings again and post results of 100 count 2.5 interval

If it's an overloading issue, it will vary as it gets used more in peak times and less outside.



Userext #14 Posted 16 October 2017 - 08:28 PM

    Warrant Officer

  • Player
  • 16259 battles
  • 897
  • Member since:
    06-12-2011

View PostBaldrickk, on 16 October 2017 - 08:21 PM, said:

If it's an overloading issue, it will vary as it gets used more in peak times and less outside.

 

My ISP is upgrading their infrastructure to be able to sustain a complete removal of speed control. But I can't put the sole blame on that upgrading/repairing because the first pic shows a decline of my ping through corebackbone and ending with server. I know their working hours and currently they aren't working on the cables.

Maybe, just maybe WG isn't at the faulty edge. Time shall show I guess.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users