Hi! I have red plug problem during the game on some popular servers. This occurs most often during change team, respaning, using phase gates and tunnels. In short, during the instant move on the map.
I made video and marked each red plug.
Hi Myx. When you red plug, it's because the server is trying to send a lot of data and it's taking too long to get to your client. Hence the choke.
1. At a first glance, this would indicate that the server bwlimit is too low (russia 2 appears to run on default settings, default bwlimit is 25kb/s (see "avg. recv." value in net_stats)).
However, most servers run okay-ish with the default settings and you appear to be the only one getting those impressive -giant- lags.
So apart from the first option, there are two more:
2. Your internet link download bandwidth is too low (unless you live rural, this one is rather unlikely)
3. Something between your client and the servers is choking the traffic.
This could be a bad ISP, bad linkup hardware/modems etc, a router thinking that traffic is udp flooding, it could also be software on your pc.
Try playing on a server that allows a high bwlimit (for example our server is currently 150kb/s instead of the default 25kb/s) and see if you still have it there.
If you do, the cause is possibly 2. or 3.
Thank you for the explanation! Can I somehow get to know the value of bwlimit on the server? Any console command? Becouse some servers works pretty well, some with red plug. The problems started a few months ago, before that all the servers worked fine.
Not really, you'd have to ask the operator. I'm contacting artem about this, but he needs to accept my invitation.
However, from the net_stats you can see tickrate (default 30), sendrate "Updates rcvd:" (default 20) and when you have choke you can see the average inbound bandwidth usage "Average rcvd:" (default 25600 bytes/second)
If a server runs 30tick, 20 sendrate and maxes out at 25kb/s, you can safely guess it's running on full default settings.
However, I have a feeling your issue is not necessarily a server configuration problem, but rather some oddity at your end.
In the video I see you spawning into a marine start void of other players, with 6 structures there and red plugging like crazy. This is unnatural.
It's almost like your client is requesting far too much data from the server than normal.
I'd like to summon some clever people who can shed some light on whether something like that is even possible.
I've joined the Russia #2 server and checked the net_stats while it was full with a round going. It does occasionally hit the bwlimit but this is 'normal' for all servers 18+ players with default config. Everything else is 5 by 5.
I have nothing even near a red plug when switching locations, even with a lot going on.
This is further indication something is amiss at your end. You say it started a few months ago. Do you remember any changes you made, for example using a different router etc.?
I made new video on Wooza server.
Note the strange messages in the console. I think your statement "client is requesting far too much data from the server" - true. Because you can see in 4:05, after I reconnected to the server, the value of average resv was above 60 kb/s. And I don't get a red plug. The problems started later.
Looks like a packet arrives at a client broken. But pingtest.net show me that i have 0% packet lost from Russia to Milan (I don't find pingtest server in Switzerland, where Wooza's server located) How i can test it more?
Another video for comparison. No red plug on another server, but some strange messages in log.
That's one of my servers and it uses default server rates except for bwlimit which is set to 100 kb/s. In spectate mode in the last minute of the video you reached 100 kb/s once.
Finding this in the netlog where he red-plugs: :-O
[831.616] 83.78.242.192 Client: Started receiving a new packet of 16516460 bytes
Those are all over the place. Choking is no surprise with that.
Regarding the pingtest, you can either ping the game server with ping command or the hop before it (find it with tracert). Or you can use advanced programs like hrping.
Let me know if you need help with that.
Since you're on a fibre internet connection, there's probably no issue there.
Hey-ho.
Somewhat not the same problem.., but;
I have a fix for ye ol' blackscreen of doom, especially for the people who get it when they join a server, 5 bucks and you get to know it.
Or just click the spoiler.
After you've launched the game, go into options, then switch to the tab that says graphics and change the line "Wait for vertical sync" to disabled, then join the server and you should have much less to almost no blackscreen-time. Afterwards you can change your V-sync to what you had it on before, apply and enjoy playing the game, but watch them lerks, dey no flap!
(Options > Graphics > V-Sync > disabled)
L-900 Lerkminator provived me with this fix and it helped me, it might help others, and that's why I post it here, since Frax is too lazy (he's not lazy, he just doesn't give a fuck) and Wooza said: "01:26 - Wooza: ja kannst posten. aber es ist nicht selbes problem wie der im thread"
I guess V-Sync should be already disabled for competitive players who don't own a 120 Hz monitor because the lag it introduces is quite perceptible in a game like NS2. Nevertheless, I have a black screen from time to time. Pretty rare though, like once in several months.
Yesterday I was playing with @Myx on YOClan.co.uk server. I was Alien and he was Marine. Myx said that he was OK (is it right, @Myx?), but I was struggling with lags! Sometimes there were messages like "SERVER Red!" or "SERVER Yellow!". Here are some lag's samples (turn on the annotations):
Also, earlier I posted a video where I got a teleport lag on YOClan server (like Myx did):
Some questions:
1. Am I experiencing lags due to exceeding the BWlimit? Then why am I exceeding it on some servers?
2. Why it was super perfect on all servers before 267 build? What was changed in 267 build?
3. Can it be related to this thread somehow?
Taking the liberty to summarize what's going on in the video:
1. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location = same thing myx has
2. server fine, upd rcvd low, avg rcvd maxed, choke 40, spawning, changing location away from spec target. that should actually still work ok. Theres just the lone gorge..
3. server fine, upd rcvd low, avg rcvd maxed briefly, choke 26, with changing location (exiting a tunnel). once again, this should normally not cause issues.
4. server fine, upd rcvd low, avg rcvd maxed briefly, choke 26, changing to spec target after dying = changing location = should work fine
5. server fine, upd rcvd low, avg rcvd maxed, choke 100, changing to spec target after dying.
6. server red, upd rcvd low, avg rcvd nominal, choke 100, this is a server issue.
6.2 server fine, upd rcvd low, avg rcvd maxed, choke 50, spawning, changing location away from spec target.
7. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location.
8. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location. (of course he does walk around, but not instantly switch from one position on the map to another far away...)
9. round end lag, disregard that.
In almost all instances you hit the bwlimit, like Myx. You're saying correctly that you're exceeding the bwlimit far too easily.
Please do a net_log 2 and check for the funny super size packets that Myx sees. And the other errors.
2nd video: same thing essentially, he phases into a room with too little going on to cause the massive spike(red plug).
Also try reproducing this on a high bwlimit server. Myx has the same thing on those.
Also try reproducing this on a high bwlimit server. Myx has the same thing on those.
Thx for comments. Today I got this on server with bwlimit 150:
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.137] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.138] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.138] MainThread : Error: OnPacket: Received an empty NetworkPacket
[2902.138] MainThread : Error: OnPacket: Received an empty NetworkPacket
with Red plug. The Red plug was there for 5 sec till I die, there were no lags tho'.
IronHorseDeveloper, QA Manager, Technical Support & contributorJoin Date: 2010-05-08Member: 71669Members, Super Administrators, Forum Admins, Forum Moderators, NS2 Developer, NS2 Playtester, Squad Five Blue, Subnautica Playtester, Subnautica PT Lead, Pistachionauts
Well that error was added in recently as a debug message from one of our engine programmers, since it was previously crashing players supposedly.
NS2 shouldn't even be doing that, though.
If those who are having these problems @abse , @METROID , @RtasVadum , @Myx
Could all list their ISP name, Router model number, modem type, Operating system, download and upload speeds, whether you're wired or wireless, and whether this happens on every server or just specific ones - it will help further diagnose this issue.
Thanks!
@Myx That tracert seems to fail in 4 different hops.
Could this be due to blocking ping requests on these points along the route?
since it was previously crashing players supposedly
I got crash few times during super lag with extra spam "OnPacket: Received an empty NetworkPacket" message in console. But I didn't give it value, because the servers were with mods.
Could all list their ISP name, Router model number, modem type, Operating system, download and upload speeds, whether you're wired or wireless, and whether this happens on every server or just specific ones - it will help further diagnose this issue.
Ufanet ISP, Routel model TP-Link WR1043ND (I tried to connect the cable directly without a router - the same result), Windows 7 Pro 64-bit, 30Mb/s download, 10Mb/s upload, wired, specific ones - at least SupaFred's servers (The Thirsty Onos) works well for me.
By the way, I tried to use VPN proxy to Netherlands, thus changing the route, and I got perfect result: no lags, no red plug. So I think it is likely a problem in the route between my ISP and the server. But i have no idea why, because pings to the servers seems ok. I tested it a lot, pings always come intact, but games packets broken.
I would definitely call your ISP and provide as much info as you can regarding where it is failing and have them reproduce the issue on the phone if you can.
I am willing to bet if you did a tracert using your VPN you would see no failed hops like you did with your normal connection.
That's on your ISP to fix , hopefully they are responsive!
p.s. pings don't mean everything, as they are not continuous and real time, they only report every few moments and are a limited query compared to a constant, live, and demanding connection that requires no packet loss. Visit www.pingtest.net to see what i mean.. there's pinging a location.. and then there's the health of the connection, like packet loss or jitter which are VERY important in gaming.
@abse i would try what i suggested previously in this thread. Run a Tracert to the same IP of a server that you have issues with and see if it fails anywhere along the way
Failed hops don't mean anything. It's just hosts that don't send ICMP replies back on TTL expiration and/or regular pings. It may be related to MTU issues or not just like any other hop. It's pretty normal to have such hops where there should be a DC or ISP gateway.
Moreover, ISP has no control over routers this far away neither it can route your traffic around them. Every single router in the Internet decides for itself where to send traffic next. Nobody can order it to route as they want except if the router admin configures it to do so. But the protocol itself doesn't allow relaying any path information.
I may suggest lowering the MTU on the network interface to, say, 1400 or even lower. There's a possibility that some host blocks relatively large packets along the way and you have no way to find it automatically with PMTU discovery. To keep things clean, connect to the ISP directly and also install Wireshark and capture the traffic. Look for the packet size, if it really changes according to the MTU setting, you're good. If it won't help, well, at least you tried.
Wooza server have new ip, so tracert them.
Left side: ping from my PC, right side - VPN
The steps are the same for 13 and 9 hops respectively. Two penultimate hops still failed. But I have no problems in game with VPN and have problems without :-/
This fail hops doesn't matter. I'm not a network specialist, but even I know that some routers can block ping requests. Even on my home router is an option! Look at this:
I trying to ping my IP by VPN
My router just block ping request! That doesnt matter.
About my ISP. I wanna tell them about my problem, but what i can tell them? "MY GAME DOESN'T WORK HELP ME!!!!" ? Support will request tracert etc. Tracerts are fine. All other games works fine. How do you think they will respond to me?
I noticed that this problems started after 267 patch. Because that I open this topic here. Maybe the problem in the route began at the same time in the release of the new patch, just a coincidence. I do not know. But I can't ask my ISP until I have no hard evidence.
pings don't mean everything, as they are not continuous and real time, they only report every few moments and are a limited query compared to a constant, live, and demanding connection that requires no packet loss. Visit www.pingtest.net to see what i mean.. there's pinging a location.. and then there's the health of the connection, like packet loss or jitter which are VERY important in gaming.
noticed Frankfurt on tracert to Wooza server. And pingtest.net also have Frankfurt server.
Another pingtest to Europe:
To Britain:
To Spain:
USA, Los Angeles:
I may suggest lowering the MTU on the network interface to, say, 1400 or even lower. There's a possibility that some host blocks relatively large packets along the way and you have no way to find it automatically with PMTU discovery. To keep things clean, connect to the ISP directly and also install Wireshark and capture the traffic. Look for the packet size, if it really changes according to the MTU setting, you're good. If it won't help, well, at least you tried.
This is strange, but with standart MTU 1492, maximum length of ns2 UDP packet was 1442. Then I changed MTU to 1380, and still get 1442 length UDP packets. WTF?
Or maybe i made mistake? Packet size in Wireshark call "length"?..
Comments
1. At a first glance, this would indicate that the server bwlimit is too low (russia 2 appears to run on default settings, default bwlimit is 25kb/s (see "avg. recv." value in net_stats)).
However, most servers run okay-ish with the default settings and you appear to be the only one getting those impressive -giant- lags.
So apart from the first option, there are two more:
2. Your internet link download bandwidth is too low (unless you live rural, this one is rather unlikely)
3. Something between your client and the servers is choking the traffic.
This could be a bad ISP, bad linkup hardware/modems etc, a router thinking that traffic is udp flooding, it could also be software on your pc.
Try playing on a server that allows a high bwlimit (for example our server is currently 150kb/s instead of the default 25kb/s) and see if you still have it there.
If you do, the cause is possibly 2. or 3.
However, from the net_stats you can see tickrate (default 30), sendrate "Updates rcvd:" (default 20) and when you have choke you can see the average inbound bandwidth usage "Average rcvd:" (default 25600 bytes/second)
If a server runs 30tick, 20 sendrate and maxes out at 25kb/s, you can safely guess it's running on full default settings.
However, I have a feeling your issue is not necessarily a server configuration problem, but rather some oddity at your end.
In the video I see you spawning into a marine start void of other players, with 6 structures there and red plugging like crazy. This is unnatural.
It's almost like your client is requesting far too much data from the server than normal.
I'd like to summon some clever people who can shed some light on whether something like that is even possible.
I have nothing even near a red plug when switching locations, even with a lot going on.
This is further indication something is amiss at your end. You say it started a few months ago. Do you remember any changes you made, for example using a different router etc.?
Note the strange messages in the console. I think your statement "client is requesting far too much data from the server" - true. Because you can see in 4:05, after I reconnected to the server, the value of average resv was above 60 kb/s. And I don't get a red plug. The problems started later.
Log.txt from %appdata%\Natural Selection 2\
https://yadi.sk/i/OmGtYAbIbwJkK
log.txt
https://yadi.sk/i/zS6DJ8eqbwKPa
That's one of my servers and it uses default server rates except for bwlimit which is set to 100 kb/s. In spectate mode in the last minute of the video you reached 100 kb/s once.
Finding this in the netlog where he red-plugs: :-O
Those are all over the place. Choking is no surprise with that.
Regarding the pingtest, you can either ping the game server with ping command or the hop before it (find it with tracert). Or you can use advanced programs like hrping.
Let me know if you need help with that.
Since you're on a fibre internet connection, there's probably no issue there.
Somewhat not the same problem.., but;
I have a fix for ye ol' blackscreen of doom, especially for the people who get it when they join a server, 5 bucks and you get to know it.
Or just click the spoiler.
After you've launched the game, go into options, then switch to the tab that says graphics and change the line "Wait for vertical sync" to disabled, then join the server and you should have much less to almost no blackscreen-time. Afterwards you can change your V-sync to what you had it on before, apply and enjoy playing the game, but watch them lerks, dey no flap!
L-900 Lerkminator provived me with this fix and it helped me, it might help others, and that's why I post it here, since Frax is too lazy (he's not lazy, he just doesn't give a fuck) and Wooza said: "01:26 - Wooza: ja kannst posten. aber es ist nicht selbes problem wie der im thread"
Also, earlier I posted a video where I got a teleport lag on YOClan server (like Myx did):
Some questions:
1. Am I experiencing lags due to exceeding the BWlimit? Then why am I exceeding it on some servers?
2. Why it was super perfect on all servers before 267 build? What was changed in 267 build?
3. Can it be related to this thread somehow?
Summoning @ATF, @Obraxis, @GhoulofGSG9
1. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location = same thing myx has
2. server fine, upd rcvd low, avg rcvd maxed, choke 40, spawning, changing location away from spec target. that should actually still work ok. Theres just the lone gorge..
3. server fine, upd rcvd low, avg rcvd maxed briefly, choke 26, with changing location (exiting a tunnel). once again, this should normally not cause issues.
4. server fine, upd rcvd low, avg rcvd maxed briefly, choke 26, changing to spec target after dying = changing location = should work fine
5. server fine, upd rcvd low, avg rcvd maxed, choke 100, changing to spec target after dying.
6. server red, upd rcvd low, avg rcvd nominal, choke 100, this is a server issue.
6.2 server fine, upd rcvd low, avg rcvd maxed, choke 50, spawning, changing location away from spec target.
7. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location.
8. server fine, upd rcvd low, avg rcvd maxed, choke 100, without changing location. (of course he does walk around, but not instantly switch from one position on the map to another far away...)
9. round end lag, disregard that.
In almost all instances you hit the bwlimit, like Myx. You're saying correctly that you're exceeding the bwlimit far too easily.
Please do a net_log 2 and check for the funny super size packets that Myx sees. And the other errors.
2nd video: same thing essentially, he phases into a room with too little going on to cause the massive spike(red plug).
Also try reproducing this on a high bwlimit server. Myx has the same thing on those.
That warrants some looking into, even if this issue ends up being about bandwidth limits.
ADSL, no fibre.
Edit:
Needless to say i'm only experiencing these lag issues with NS2.
Everything else works perfectly.
NS2 shouldn't even be doing that, though.
If those who are having these problems @abse , @METROID , @RtasVadum , @Myx
Could all list their ISP name, Router model number, modem type, Operating system, download and upload speeds, whether you're wired or wireless, and whether this happens on every server or just specific ones - it will help further diagnose this issue.
Thanks!
By the way, I tried to use VPN proxy to Netherlands, thus changing the route, and I got perfect result: no lags, no red plug. So I think it is likely a problem in the route between my ISP and the server. But i have no idea why, because pings to the servers seems ok. I tested it a lot, pings always come intact, but games packets broken.
Windows 7 64 bit
Wired
It happens in every server i tried, so far
I would definitely call your ISP and provide as much info as you can regarding where it is failing and have them reproduce the issue on the phone if you can.
I am willing to bet if you did a tracert using your VPN you would see no failed hops like you did with your normal connection.
That's on your ISP to fix , hopefully they are responsive!
p.s. pings don't mean everything, as they are not continuous and real time, they only report every few moments and are a limited query compared to a constant, live, and demanding connection that requires no packet loss. Visit www.pingtest.net to see what i mean.. there's pinging a location.. and then there's the health of the connection, like packet loss or jitter which are VERY important in gaming.
@abse i would try what i suggested previously in this thread. Run a Tracert to the same IP of a server that you have issues with and see if it fails anywhere along the way
Moreover, ISP has no control over routers this far away neither it can route your traffic around them. Every single router in the Internet decides for itself where to send traffic next. Nobody can order it to route as they want except if the router admin configures it to do so. But the protocol itself doesn't allow relaying any path information.
I may suggest lowering the MTU on the network interface to, say, 1400 or even lower. There's a possibility that some host blocks relatively large packets along the way and you have no way to find it automatically with PMTU discovery. To keep things clean, connect to the ISP directly and also install Wireshark and capture the traffic. Look for the packet size, if it really changes according to the MTU setting, you're good. If it won't help, well, at least you tried.
Left side: ping from my PC, right side - VPN
The steps are the same for 13 and 9 hops respectively. Two penultimate hops still failed. But I have no problems in game with VPN and have problems without :-/
This fail hops doesn't matter. I'm not a network specialist, but even I know that some routers can block ping requests. Even on my home router is an option! Look at this:
I trying to ping my IP by VPN
My router just block ping request! That doesnt matter.
About my ISP. I wanna tell them about my problem, but what i can tell them? "MY GAME DOESN'T WORK HELP ME!!!!" ? Support will request tracert etc. Tracerts are fine. All other games works fine. How do you think they will respond to me?
I noticed that this problems started after 267 patch. Because that I open this topic here. Maybe the problem in the route began at the same time in the release of the new patch, just a coincidence. I do not know. But I can't ask my ISP until I have no hard evidence.
http://forums.unknownworlds.com/discussion/comment/2220335/#Comment_2220335
Another pingtest to Europe:
To Britain:
To Spain:
USA, Los Angeles:
Where problems with my ISP?
Or maybe i made mistake? Packet size in Wireshark call "length"?..