Shooting ahead

nämeless · 2012-12-05 20:59:31

Thanks for everyone, the problem is solved by a practical experiment. I have discovered one thing.
1. When I ping the server from Windows console - no packet loss as well.
2. When I ping the server from Windows console with options -l 1500, where 1500 bytes is a standard of MTU size for Ethernet that Windows can set by default, I got packet loss about 50% but with the keys -l 1500 and -f that disallows the packet fragmentation, I got the message "Fragmentation required".

I have pinged some servers to get randomly optimal value without any errors, changed the MTU size and disabled TCP auto-tuning. Everything is okay now.

Last edited by nämeless (2012-12-05 21:16:45)

joint · 2012-12-05 22:54:22

so you saying that the value of MTU preferably has to be multiple of ping? and that what a "bad ping" about?
i mean for example if your ping is 50ms, you need to know how many MTU blocks will pass in your connection while 50ms, and make such MTU's size, that division on ping without residue, or small residue, because this residues happens just losing(надеюсь слово residue-это из математики "остаток")
i bet i typed unreadable xD

s[sk] · 2012-12-05 23:05:35

nämeless wrote:

Thanks for everyone, the problem is solved by a practical experiment. I have discovered one thing.
1. When I ping the server from Windows console - no packet loss as well.
2. When I ping the server from Windows console with options -l 1500, where 1500 bytes is a standard of MTU size for Ethernet that Windows can set by default, I got packet loss about 50% but with the keys -l 1500 and -f that disallows the packet fragmentation, I got the message "Fragmentation required".
I have pinged some servers to get randomly optimal value without any errors, changed the MTU size and disabled TCP auto-tuning. Everything is okay now.

i'm happy that you've solved you problem, i just want to point out that your methodology was not completely correct

1. it's udp, so tcp auto tuning has no impact on this
2. icmp header has 8 bytes, ip header has 20 bytes so the max you get with mtu 1500 and no fragmentation is 1500-28=1472 bytes

so if you actually had problems with ping sizes < 1472 with mtu 1500, then you'd set mtu to "ok size" + 28

just in case anyone else tries this too

EDIT: actually i'm not happy that you've fixed it because now you'll be killing me even more...

Last edited by s[sk] (2012-12-05 23:09:22)

joint · 2012-12-05 23:44:24

s[sk] how do you think, is possible that the integrity of number of net frames while synchronization time (ping) may influence on ingame hits calculation, or this is completely nonsense?

nämeless · 2012-12-06 01:25:39

s[sk] wrote:

1. it's udp, so tcp auto tuning has no impact on this

I know but the interesting moment is that without disabled auto-tuning it doesn't work. I don't know how to explain that.

netsh int tcp set global autotuninglevel=disabled

s[sk] wrote:

so if you actually had problems with ping sizes < 1472 with mtu 1500, then you'd set mtu to "ok size" + 28

No, I think differently from you and I use a value of 1460, subtracting 20 + 20 bytes of IP and TCP headers.
UPD. Fuck, you don't let me go to sleep. After googling, I understood that I was incorrect cause I forgot that I using VPN L2TP connection. So, it can be an explanation:

IP header = 20 bytes
UDP header = 8 bytes
L2TP header = 12 bytes
the version and flag fields (2 bytes)
the tunnel id and session id fields (2 bytes each)
2 bytes of padding offset
4 bytes of Point-to-Point Protocol (PPP) encapsulation

1500 - 20 - 8 - 12 = 1460;

joint wrote:

so you saying that the value of MTU preferably has to be multiple of ping? and that what a "bad ping" about?

For me, it doesn't impact on ping value but it impacts on packet loss only to this server. Who knows, maybe 1500 bytes is too thick to fit into the graceful and elegant French routing.

s[sk] · 2012-12-06 01:28:14

nämeless wrote:

UPD. Fuck, you don't let me go to sleep. After googling, I understood that I was incorrect cause I forgot that I using VPN L2TP connection.

yes, that explains a lot
and now tuia should ban you for hiding yourself behind vpn!

j/k ;-)

s[sk] · 2012-12-06 01:31:23

joint wrote:

s[sk] how do you think, is possible that the integrity of number of net frames while synchronization time (ping) may influence on ingame hits calculation, or this is completely nonsense?

sorry but i really have no idea what you're asking
try russian, nameless will translate it for me ;-)

nämeless · 2012-12-06 01:43:32

s[sk] wrote:

joint wrote:
s[sk] how do you think, is possible that the integrity of number of net frames while synchronization time (ping) may influence on ingame hits calculation, or this is completely nonsense?
sorry but i really have no idea what you're asking
try russian, nameless will translate it for me ;-)

Oh, as far as I know, he is a Cisco engineer. Probably, throughout the years of weed smoking during the working with Cisco hardware via PuTTy console makes his language gibberish and "jointish".

Last edited by nämeless (2012-12-06 01:52:05)

joint · 2012-12-06 02:45:34

nämeless wrote:

s[sk] wrote:
joint wrote:
s[sk] how do you think, is possible that the integrity of number of net frames while synchronization time (ping) may influence on ingame hits calculation, or this is completely nonsense?
sorry but i really have no idea what you're asking
try russian, nameless will translate it for me ;-)
Oh, as far as I know, he is a Cisco engineer. Probably, throughout the years of weed smoking during the working with Cisco hardware via PuTTy console makes his language gibberish and "jointish".

no i just start the main course
i will try again
for example i have ping 50 ms, its 0.05 sec, between the moment when i click "fire" and the moment when server receiving my data runing out-0.025 sec, that mean during 0.025 sec a certain number of packets is passed from client to server,
every packet its like a atom, a quantum of communication, right? and the atom of the synchronization- its ping time
in every 0.025 some of packets may will "not shape" in the time needed to send data to server and receive data back ,this may cause some changes..
so the question: how many weed i had to smoked to wrote this?

Last edited by joint (2012-12-06 02:51:46)

nämeless · 2012-12-06 10:16:31

joint wrote:

every packet its like a atom, a quantum of communication, right? and the atom of the synchronization- its ping time
in every 0.025 some of packets may will "not shape" in the time needed to send data to server and receive data back ,this may cause some changes..
so the question: how many weed i had to smoked to wrote this?

How does marijuana affect the brain?

joint · 2012-12-06 15:12:25

nämeless wrote:

How does marijuana affect the brain?

dont ask, if i tell you - you will want to try it, then you become a drug addict and will start rob some schoolkids taking their cellphones away on the streets
I could not live with it, sorry

Timmos · 2012-12-06 15:41:53

check holland, we got robbers and muglers all over the street asking for a joint

tuia · 2012-12-06 16:27:27

joint wrote:

i will try again
for example i have ping 50 ms, its 0.05 sec, between the moment when i click "fire" and the moment when server receiving my data runing out-0.025 sec, that mean during 0.025 sec a certain number of packets is passed from client to server,
every packet its like a atom, a quantum of communication, right? and the atom of the synchronization- its ping time
in every 0.025 some of packets may will "not shape" in the time needed to send data to server and receive data back ,this may cause some changes..
so the question: how many weed i had to smoked to wrote this?

Your UDP packets take 50 ms to reach the server, and the packets from the server to you take another 50 ms. There is no lag compensation, so you have to shoot ahead in time to hit according to your ping, I'm sure you know it.

joint · 2012-12-06 18:44:26

lol i have no idea where that 0.025 come from 8|

nämeless · 2012-12-30 23:59:48

The problem is still exists.

Трассировка маршрута к ks212581.kimsufi.com [188.165.220.115]
с максимальным числом прыжков 30:

  1     1 ms     1 ms     1 ms  asr-1-nnov.kis.ru [92.242.92.245]
  2     1 ms     2 ms     1 ms  195.98.51.17
  3     1 ms     1 ms     1 ms  195.239.15.53
  4     *        *        *     Превышен интервал ожидания для запроса.
  5    71 ms    70 ms    70 ms  rbx-g2-a9.fr.eu [91.121.131.130]
  6    72 ms     *        *     vss-4-6k.fr.eu [91.121.131.166]
  7    62 ms    62 ms    62 ms  ks212581.kimsufi.com [188.165.220.115]

Трассировка завершена.

s[sk] · 2012-12-31 00:07:21

it's normal that routers especially core ones drop icmp packets targeted to them because they consider them low priority, you can ignore that

nämeless · 2012-12-31 00:41:36

Okay, Windows tracert uses ICMP. I will try to install Linux to make traceroute with UDP packets.

nämeless · 2012-12-31 01:21:47

Thanks. I just have installed PingPlotter to set UDP Packets in options but seems to me it doesn't work.

nämeless · 2013-03-26 16:48:00

The problem is getting back. Today's video of dogfights with Bio on El Alamein as an example. It's also a bit laggy. Watch in 1080p HD to understand what I mean.

http://www.youtube.com/watch?v=U6OEN3ddRIQ

Last edited by nämeless (2013-03-26 16:57:46)

s[sk] · 2013-03-26 17:26:56

you've cut the part where i shot you down! bastard!

nämeless · 2013-03-26 17:32:36

I remember that moment: I was from behind of enemy Spitfire and you were from behind of me. I didn't cut this part because I didn't record. I use Fraps and made only a few moments because of huge size of lossless Full HD video.

P. S. I have bought a joystick a couple of days earlier and I still don't get used to it. I think my old keyboard is the way better.

Last edited by nämeless (2013-03-26 17:39:55)

Timmos · 2013-03-26 19:25:45

nameless, ive watched the video for 2 minutes and its nothing with ur ping.. at first i see some aiming that isnt good, and second.ask any pilot about the no hits on ivo

btw now i also see why i can easy win u in df, practice more with ur views!

joint · 2013-03-26 19:41:02

DF TOURNAMENT!!!!

dary · 2013-03-26 19:50:41

hits look normal to me. one time, i don't know which minute, you aimed too much in front, because he was pretty slow you didn't hit at all.

Ivo knows how to escape pretty good, i mean it is not easy to find the right spot to hit him sometimes. are you sure this is not mato?

Last edited by dary (2013-03-26 19:53:58)

s[sk] · 2013-03-26 20:47:32

Timmos wrote:

btw now i also see why i can easy win u in df, practice more with ur views!

more? i've had an epileptic seizure while watching it!

Timmos · 2013-03-27 18:45:21

s[sk] wrote:

Timmos wrote:
btw now i also see why i can easy win u in df, practice more with ur views!
more? i've had an epileptic seizure while watching it!

he used them wrong,

Arkos · 2014-06-06 19:27:33

...

Hi all...

I guess, everybody from you reminds my "whining" about lagy feeling on your server. Sometimes i can hit an opponent which is standing still 8 times directly into head, and nothing happends. Sometimes, while flying, when im directly behind another plane, i need 400 bullet's to take it down. In fights, i have to aim, like i would have 200-400 ping, but even then, when bullet's are hitting, some of them have no effect.

I could live with that. But the really worst thing is, the situation changes allways while playing. That means, i have allways to readjust my aimings, even in the same map, but in droping bombs the differences are sometimes to sick.

Today i made a short test.

Is this the reason for these problems?

...

(PS: And don't blame me, (again), for my english, i can't speak nor write in english...)

...

Last edited by Arkos (2014-06-06 20:16:20)

bud · 2014-06-06 19:59:12

Looks like you playing through a vpn, if you are you cant see the vpn routing with HLSW

Arkos · 2014-06-06 20:13:04

no vpn

Denny · 2014-06-06 21:06:14

hey arkos

SiMPLE | Forum

Announcement

#31 2012-12-05 20:59:31

Re: Shooting ahead

#32 2012-12-05 22:54:22

Re: Shooting ahead

#33 2012-12-05 23:05:35

Re: Shooting ahead

#34 2012-12-05 23:44:24

Re: Shooting ahead

#35 2012-12-06 01:25:39

Re: Shooting ahead

#36 2012-12-06 01:28:14

Re: Shooting ahead

#37 2012-12-06 01:31:23

Re: Shooting ahead

#38 2012-12-06 01:43:32

Re: Shooting ahead

#39 2012-12-06 02:45:34

Re: Shooting ahead

#40 2012-12-06 10:16:31

Re: Shooting ahead

#41 2012-12-06 15:12:25

Re: Shooting ahead

#42 2012-12-06 15:41:53

Re: Shooting ahead

#43 2012-12-06 16:27:27

Re: Shooting ahead

#44 2012-12-06 18:44:26

Re: Shooting ahead

#45 2012-12-30 23:59:48

Re: Shooting ahead

#46 2012-12-31 00:07:21

Re: Shooting ahead

#47 2012-12-31 00:41:36

Re: Shooting ahead

#48 2012-12-31 01:21:47

Re: Shooting ahead

#49 2013-03-26 16:48:00

Re: Shooting ahead

#50 2013-03-26 17:26:56

Re: Shooting ahead

#51 2013-03-26 17:32:36

Re: Shooting ahead

#52 2013-03-26 19:25:45

Re: Shooting ahead

#53 2013-03-26 19:41:02

Re: Shooting ahead

#54 2013-03-26 19:50:41

Re: Shooting ahead

#55 2013-03-26 20:47:32

Re: Shooting ahead

#56 2013-03-27 18:45:21

Re: Shooting ahead

#57 2014-06-06 19:27:33

Re: Shooting ahead

#58 2014-06-06 19:59:12

Re: Shooting ahead

#59 2014-06-06 20:13:04

Re: Shooting ahead

#60 2014-06-06 21:06:14

Re: Shooting ahead

Board footer