Darn! Connection Closed!

[panzer-runner 2]

Mokpama_rapra and Punzer-runner

You are such experienced programmers (sarcasm) ... Thank you very much for your advise. Before that we really did not know how to work! We were such a noobs! (double sarcasm)

Don't teach us how to work.

Roman Epishin wrote:

Dear players, due to work on the optimization of the project may cause slight disruption to the servers, leading to the so-called "lags". We apologize for this inconvenience and ask for patience. We know about all the pitfalls, and the team "Tanks Online operates in emergency mode to complete the optimization work in the shortest possible time. Thank you for being with us!

 

Rude as usual. Thanks for not being helpful, but taking time to be rude and obnoxious. You are a programmer? You are a developer here? Really? I thought you were a cop! :roll: :twisted: What I find funny is that the real devs here in your post ADMIT they have problems, and are / did work on it. The whole time we were / are being told by others like yourself that the problem is NOT with Tanki, but is instead with the player. FAIL.

 

Quick side question for anyone who can answer: When one player gets a lag due to internet connectivity problems' date=' do all the other players in the battle get the lag too? What is the cause of lags?[/quote']

In short, no (if other players do not have the same problem with their connections).

Now, when server lags are virtually gone, the most popular are microlags due to packet loss (somewhere on the path between a player and a tanki server). Their symptoms are:

 

 a) the lagging player sees himself moving smoothly, but other players are twitching for him;

 b) other players see each other normally, but the lagging player is twitching for them.

 

There is a quick and reliable way to check packet loss on your path to tanki servers: in the command prompt, type ping -n 100 c5.tankionline.com (you can use your favorite server number instead of "5"), wait until it finishes (100 pings), then look at the statistics: normally, there should be 0% loss, 1 or 2 would be satisfactory, more than 3 would lead to uncomfortable gameplay and unreliable connection.

 

As for "connection closed by server" – if you have this consistently when you log in, you may be blocked and should contact the technical support team. Otherwise, you probably have some kind of connectivity problem. Read this post. Excessive packet loss also may be a cause for this.

 

A separate message, "connection to server failed", generally indicates that something (probably firewall) blocks outgoing connections to port 5222.

 

As usual, no one blames Tanki for anything. It is ALWAYS the fault of the player. :roll:

 

For weeks the lag problem was fixed, with virtually NO problems. Recently it is back worse than ever. It very well could be the bandwidth on the path to the servers, but then again, Tanki should not have all servers in one location. That is totally UNBALANCED for traffic load. As for it being packet loss causing the problem, I can show otherwise. I just tried to play on various servers and games all with horrible lag in game , unplayable.

 

I also immediately ran a ping test to the server I was on last with problems.

 

PING c1.tankionline.com (62.152.57.11): 56 data bytes

64 bytes from 62.152.57.11: icmp_seq=0 ttl=49 time=169.032 ms

64 bytes from 62.152.57.11: icmp_seq=1 ttl=49 time=171.507 ms

64 bytes from 62.152.57.11: icmp_seq=2 ttl=49 time=169.429 ms

64 bytes from 62.152.57.11: icmp_seq=3 ttl=49 time=176.356 ms

64 bytes from 62.152.57.11: icmp_seq=4 ttl=49 time=172.318 ms

64 bytes from 62.152.57.11: icmp_seq=5 ttl=49 time=170.356 ms

64 bytes from 62.152.57.11: icmp_seq=6 ttl=49 time=170.708 ms

64 bytes from 62.152.57.11: icmp_seq=7 ttl=49 time=171.873 ms

64 bytes from 62.152.57.11: icmp_seq=8 ttl=49 time=168.456 ms

64 bytes from 62.152.57.11: icmp_seq=9 ttl=49 time=176.977 ms

64 bytes from 62.152.57.11: icmp_seq=11 ttl=49 time=167.906 ms

64 bytes from 62.152.57.11: icmp_seq=13 ttl=49 time=169.365 ms

64 bytes from 62.152.57.11: icmp_seq=14 ttl=49 time=170.648 ms

64 bytes from 62.152.57.11: icmp_seq=15 ttl=49 time=172.372 ms

64 bytes from 62.152.57.11: icmp_seq=16 ttl=49 time=176.458 ms

64 bytes from 62.152.57.11: icmp_seq=18 ttl=49 time=168.322 ms

64 bytes from 62.152.57.11: icmp_seq=19 ttl=49 time=169.290 ms

64 bytes from 62.152.57.11: icmp_seq=20 ttl=49 time=168.834 ms

64 bytes from 62.152.57.11: icmp_seq=21 ttl=49 time=167.513 ms

64 bytes from 62.152.57.11: icmp_seq=22 ttl=49 time=175.399 ms

64 bytes from 62.152.57.11: icmp_seq=23 ttl=49 time=172.245 ms

64 bytes from 62.152.57.11: icmp_seq=24 ttl=49 time=170.703 ms

64 bytes from 62.152.57.11: icmp_seq=25 ttl=49 time=171.266 ms

64 bytes from 62.152.57.11: icmp_seq=26 ttl=49 time=169.293 ms

64 bytes from 62.152.57.11: icmp_seq=27 ttl=49 time=169.131 ms

64 bytes from 62.152.57.11: icmp_seq=28 ttl=49 time=170.618 ms

64 bytes from 62.152.57.11: icmp_seq=29 ttl=49 time=169.130 ms

64 bytes from 62.152.57.11: icmp_seq=30 ttl=49 time=168.383 ms

64 bytes from 62.152.57.11: icmp_seq=31 ttl=49 time=167.960 ms

64 bytes from 62.152.57.11: icmp_seq=33 ttl=49 time=171.188 ms

64 bytes from 62.152.57.11: icmp_seq=34 ttl=49 time=172.186 ms

64 bytes from 62.152.57.11: icmp_seq=35 ttl=49 time=168.790 ms

64 bytes from 62.152.57.11: icmp_seq=36 ttl=49 time=174.644 ms

64 bytes from 62.152.57.11: icmp_seq=37 ttl=49 time=167.784 ms

64 bytes from 62.152.57.11: icmp_seq=38 ttl=49 time=169.276 ms

64 bytes from 62.152.57.11: icmp_seq=39 ttl=49 time=171.422 ms

64 bytes from 62.152.57.11: icmp_seq=40 ttl=49 time=168.280 ms

64 bytes from 62.152.57.11: icmp_seq=41 ttl=49 time=168.445 ms

64 bytes from 62.152.57.11: icmp_seq=42 ttl=49 time=168.414 ms

64 bytes from 62.152.57.11: icmp_seq=43 ttl=49 time=169.340 ms

64 bytes from 62.152.57.11: icmp_seq=44 ttl=49 time=166.996 ms

64 bytes from 62.152.57.11: icmp_seq=46 ttl=49 time=169.832 ms

64 bytes from 62.152.57.11: icmp_seq=47 ttl=49 time=172.291 ms

64 bytes from 62.152.57.11: icmp_seq=48 ttl=49 time=168.521 ms

64 bytes from 62.152.57.11: icmp_seq=49 ttl=49 time=170.605 ms

64 bytes from 62.152.57.11: icmp_seq=50 ttl=49 time=169.330 ms

64 bytes from 62.152.57.11: icmp_seq=51 ttl=49 time=168.334 ms

64 bytes from 62.152.57.11: icmp_seq=52 ttl=49 time=169.031 ms

64 bytes from 62.152.57.11: icmp_seq=53 ttl=49 time=166.479 ms

64 bytes from 62.152.57.11: icmp_seq=55 ttl=49 time=169.152 ms

64 bytes from 62.152.57.11: icmp_seq=56 ttl=49 time=169.001 ms

64 bytes from 62.152.57.11: icmp_seq=57 ttl=49 time=177.935 ms

64 bytes from 62.152.57.11: icmp_seq=59 ttl=49 time=166.147 ms

64 bytes from 62.152.57.11: icmp_seq=60 ttl=49 time=167.937 ms

64 bytes from 62.152.57.11: icmp_seq=62 ttl=49 time=168.654 ms

64 bytes from 62.152.57.11: icmp_seq=63 ttl=49 time=172.020 ms

64 bytes from 62.152.57.11: icmp_seq=64 ttl=49 time=168.704 ms

64 bytes from 62.152.57.11: icmp_seq=65 ttl=49 time=169.468 ms

64 bytes from 62.152.57.11: icmp_seq=66 ttl=49 time=194.522 ms

64 bytes from 62.152.57.11: icmp_seq=67 ttl=49 time=169.095 ms

64 bytes from 62.152.57.11: icmp_seq=68 ttl=49 time=171.254 ms

64 bytes from 62.152.57.11: icmp_seq=70 ttl=49 time=168.302 ms

64 bytes from 62.152.57.11: icmp_seq=72 ttl=49 time=168.938 ms

64 bytes from 62.152.57.11: icmp_seq=73 ttl=49 time=178.320 ms

64 bytes from 62.152.57.11: icmp_seq=74 ttl=49 time=168.647 ms

64 bytes from 62.152.57.11: icmp_seq=75 ttl=49 time=167.579 ms

64 bytes from 62.152.57.11: icmp_seq=76 ttl=49 time=173.228 ms

64 bytes from 62.152.57.11: icmp_seq=77 ttl=49 time=170.900 ms

64 bytes from 62.152.57.11: icmp_seq=79 ttl=49 time=169.800 ms

64 bytes from 62.152.57.11: icmp_seq=80 ttl=49 time=172.814 ms

64 bytes from 62.152.57.11: icmp_seq=81 ttl=49 time=169.615 ms

64 bytes from 62.152.57.11: icmp_seq=82 ttl=49 time=169.049 ms

64 bytes from 62.152.57.11: icmp_seq=83 ttl=49 time=168.383 ms

64 bytes from 62.152.57.11: icmp_seq=84 ttl=49 time=173.147 ms

64 bytes from 62.152.57.11: icmp_seq=85 ttl=49 time=169.184 ms

64 bytes from 62.152.57.11: icmp_seq=86 ttl=49 time=168.497 ms

64 bytes from 62.152.57.11: icmp_seq=87 ttl=49 time=168.706 ms

64 bytes from 62.152.57.11: icmp_seq=88 ttl=49 time=181.222 ms

64 bytes from 62.152.57.11: icmp_seq=89 ttl=49 time=166.410 ms

64 bytes from 62.152.57.11: icmp_seq=90 ttl=49 time=166.822 ms

64 bytes from 62.152.57.11: icmp_seq=91 ttl=49 time=170.318 ms

64 bytes from 62.152.57.11: icmp_seq=92 ttl=49 time=169.758 ms

64 bytes from 62.152.57.11: icmp_seq=93 ttl=49 time=168.878 ms

64 bytes from 62.152.57.11: icmp_seq=94 ttl=49 time=169.561 ms

64 bytes from 62.152.57.11: icmp_seq=95 ttl=49 time=176.577 ms

64 bytes from 62.152.57.11: icmp_seq=96 ttl=49 time=167.677 ms

64 bytes from 62.152.57.11: icmp_seq=97 ttl=49 time=166.913 ms

64 bytes from 62.152.57.11: icmp_seq=98 ttl=49 time=168.766 ms

64 bytes from 62.152.57.11: icmp_seq=99 ttl=49 time=171.505 ms

 

--- c1.tankionline.com ping statistics ---

100 packets transmitted, 89 packets received, 11% packet loss

round-trip min/avg/max/stddev = 166.147/170.522/194.522/3.813 ms

 

If Tanki can't handle 11% average packet loss and a ping of around 169 without experiencing laggy games, there is something wrong with it's "look ahead" algorythm , not our connections. The server load is far more likely than traffic quality, or traffic load.

[Ippi 0]

There is a quick and reliable way to check packet loss on your path to tanki servers: in the command prompt' date=' type ping -n 100 c5.tankionline.com (you can use your favorite server number instead of "5"), wait until it finishes (100 pings), then look at the statistics: normally, there should be 0% loss, 1 or 2 would be satisfactory, more than 3 would lead to uncomfortable gameplay and unreliable connection.

 

As for "connection closed by server" – if you have this consistently when you log in, you may be blocked and should contact the technical support team. Otherwise, you probably have some kind of connectivity problem. Read this post. Excessive packet loss also may be a cause for this.

 

A separate message, "connection to server failed", generally indicates that something (probably firewall) blocks outgoing connections to port 5222.

 

As usual, no one blames Tanki for anything. It is ALWAYS the fault of the player. :roll:

I'm not blaming you, I just was trying to help you.

 

For weeks the lag problem was fixed, with virtually NO problems. Recently it is back worse than ever. It very well could be the bandwidth on the path to the servers, but then again, Tanki should not have all servers in one location. That is totally UNBALANCED for traffic load. As for it being packet loss causing the problem, I can show otherwise. I just tried to play on various servers and games all with horrible lag in game , unplayable.

 

I also immediately ran a ping test to the server I was on last with problems.

 

--- c1.tankionline.com ping statistics ---

100 packets transmitted, 89 packets received, 11% packet loss

round-trip min/avg/max/stddev = 166.147/170.522/194.522/3.813 ms

 

If Tanki can't handle 11% average packet loss and a ping of around 169 without experiencing laggy games, there is something wrong with it's "look ahead" algorythm , not our connections. The server load is far more likely than traffic quality, or traffic load.

Believe me, 11% is quite a lot. Your average ping is also close to critical, but still acceptable.

And the problem is not in server load since players in Russia and neighboring countries generally do not suffer from this. I personally play via a long Internet path through Ukraine, Poland, Germany, [some more countries], and finally Russia – still having 0% loss. You can use tracert or one of it's more advanced versions to find out where is the bottleneck.

 

Please note that today and yesterday there were problems on one of the backbone providers, that sort of things happens from time to time.

 

Well, tanki are somewhat sensitive to connection quality (probably due to Flash implementation), but they are not requiring something extraordinary.

 

If your problem is permanent, you can try to minimize the impact of packet loss with lowering MTU to the reasonable minimum (576) and disabling Nagle's algorithm (if you have Windows 7 or Vista) as described here. Don't expect a miracle though.

[r_AdvocatusDiaboli0 861]

Wow dude, 11% Packet loss! Check those wired connections in your home!

 

I'm in Pakistan and I have 0% Packet Loss. =P

[Guest CaptainFlipz]

My problem is that whenever I am in a large battle and in line to get a LOT of crystals from being in second or third place, the server decides to crash. EVERY. FREAKING. TIME. And then when I try to go back to the server, it won't load...but all the others will. I know it sounds whiny, but is there some way Tanki could take a "snapshot" of the player list for all long matches* on a server every fifteen (15) minutes or so? That way, if the server goes down, when it comes back it could give the players something instead of them having wasted four hours on a match that crashes.

 

 

* I would define a "long match" as one that lasts longer than 30 minutes.

[r_St.Evergreen0 3]

Usually the longest matches would be a 999 flag or 999 minutes match. But not a lot of people can endure to play for half a day.

[XxSdarknessSxX 2]

Or try Google Chrome.

 

This is what i use to play Tankionline now.

[Ippi 0]

My problem is that whenever I am in a large battle and in line to get a LOT of crystals from being in second or third place' date=' the server decides to crash. EVERY. FREAKING. TIME. And then when I try to go back to the server, it won't load...but all the others will. I know it sounds whiny, but is there some way Tanki could take a "snapshot" of the player list for all long matches* on a server every fifteen (15) minutes or so? That way, if the server goes down, when it comes back it could give the players [b']something[/b] instead of them having wasted four hours on a match that crashes.

Tanki do start making a snapshot every minute for a battle when its fund reaches 200 crystals. So, if the server had crashed indeed (not sure it's the case), you can request a compensation.

[king_beast88 19]

who do send request to out of curiosity? helptanki or an admin?

[Ippi 0]

who do send request to out of curiosity? helptanki or an admin?

help@tankionline.com, providing an informative message subject and detailed description of problem.

[Guest CaptainFlipz]

My problem is that whenever I am in a large battle and in line to get a LOT of crystals from being in second or third place' date=' the server decides to crash. EVERY. FREAKING. TIME. And then when I try to go back to the server, it won't load...but all the others will. I know it sounds whiny, but is there some way Tanki could take a "snapshot" of the player list for all long matches* on a server every fifteen (15) minutes or so? That way, if the server goes down, when it comes back it could give the players [b']something[/b] instead of them having wasted four hours on a match that crashes.

Tanki do start making a snapshot every minute for a battle when its fund reaches 200 crystals. So, if the server had crashed indeed (not sure it's the case), you can request a compensation.

 

Thank you! And just FYI, it looks like a compensation was added to me at some point since I last logged in. :mrgreen:

[chriswu 25]

@panzer: You must be a great writer.