Improvements for the dedicated server

[Alloc]

Steam login for web interface seems to be broken.
Steps to reproduce:

• Click the log in to steam.
  
• Sent to '/session/login' and immediately redirected to steamcommunity site.
  
• I login to steamcommunnity and get sent back to my server url at '/session/verify?all the open id stuffs here'
  
• Only a 100% empty html page is loaded.
  

The console then has this error

Other than a temporary OpenID system failure on Steam's end yesterday there's been no issues (and also no code change in a long time). Must be something specific on your end. Does the issue persist? What's the full log, at least around the error?

Searching through the log reveals: "Steam certificate error: RemoteCertificateNameMismatch, RemoteCertificateChainErrors" several lines remove from the other.

The RemoteCertificateChainErrors is ok, it's just because the game server doesn't have the root cert that's responsible for Steam's cert and thus can't validate it. That's "ok" though and shows for everyone.

RemoteCertificateNameMismatch shouldn't be in there though. Almost sounds like something is making you connect to another server or things like that.

Hey, I think I have a problem with alloc fixes ..The server is running for a while but then the server crashes.

Plain old out of memory crash, not related to my mod.

in addition, I have just found that..

[s_API FAIL] SteamAPI_Init() failed; SteamAPI_IsSteamRunning() failed.
Setting breakpad minidump AppID = 251570

Native stacktrace:

/home/sdtd2/serverfiles/7DaysToDieServer_Data/Mono/x86/libmono.so(+0x8960f) [0xf6d1f60f]
/home/sdtd2/serverfiles/7DaysToDieServer_Data/Mono/x86/libmono.so(+0x21a20) [0xf6cb7a20]
linux-gate.so.1(__kernel_rt_sigreturn+0) [0xf7731cb0]
/lib/i386-linux-gnu/libc.so.6(+0x7782c) [0xf73bf82c]
/lib/i386-linux-gnu/libc.so.6(+0x69a1d) [0xf73b1a1d]
/lib/i386-linux-gnu/libc.so.6(fread+0x77) [0xf73a6487]
./7DaysToDieServer.x86() [0x850e247]
./7DaysToDieServer.x86() [0x84eb89e]
./7DaysToDieServer.x86() [0x850ceea]
./7DaysToDieServer.x86() [0x852f899]
./7DaysToDieServer.x86() [0x853008f]
./7DaysToDieServer.x86() [0x84e47c8]
/lib/i386-linux-gnu/libpthread.so.0(+0x627a) [0xf76fc27a]
/lib/i386-linux-gnu/libc.so.6(clone+0x66) [0xf742fb56]

this always happens when the server is restarted...

You mean it's in there when you issued an shutdown? Note that the stacktraces on Unity's Linux builds are always written to the beginning of the log, not the place where the crash actually happens.

yesterday I also had an error message from allocs mod where the command "lp - list players" was executed.

Without the actual error nothing can be done here.

 

[Devidian]

Plain old out of memory crash, not related to my mod.

You should mention that it's a bug in the currently used version of unity not using more than 16G of RAM.

@Hagalaz weren't u using my scripts ? That is exactly what my out of memory script was for.

 

[Alloc]

With 4 GiB being used? Doubt so

 

[varimathras]

Ingame Map Reset On Every Restart

Hi alloc, nice to meet you.

Trekkan told me to ask you for this issue if you can help me or know something.

Im having a strange problem with the server that never had before, I notice that everytime mi modem change my ip or the server restart, the ingame map appears as undiscovered for every survival, the map was almost 75% discovered I saw it in the web map of the manager that also shows me this message every time I use it but I press Yes and opens normally. Another strange thing is that with every restart when I enter in my server it began to download hub cell dont know why.

View attachment 23931

Apart from that I dont know why the map of everyone show undiscovered like they never walked in the map, but in the web map shows what the player discovered but ingame(M) is like they are new in the map like its their first time o the map for them was reset.

This already happened sometimes, Im already using a program for a dynamic ip but I restarted the server because a player was dupping and I notice it because the server was eating almost 1gb extra of the normal

And finally an error that maybe be a glitch but dont know it appears that in some point after a certain days the server is like it stucks in a single day, in the last map we were about day 360, suddenly one day a made a quick restart and the server went back to day 348, this happened everything I restarted the server.

It looks like the server at some point stuck in certain day and rollback to that everytime you restart it.

I restarted a lot of time before the server and didnt rollback but after that day it started to happen

 

[Alloc]

I notice that everytime mi modem change my ip or the server restart, the ingame map appears as undiscovered

The uncovered map areas ingame on clients are only bound to the server's world GUID. This means that it should work as long as your server properly saves it's world header file (main.ttw).

that also shows me this message every time I use it but I press Yes and opens normally.

Can't help you with that, that must be something within RAT's browser engine.

Another strange thing is that with every restart when I enter in my server it began to download hub cell dont know why.

Same as with the uncovered maps for clients: Bound to the world GUID. Not properly getting the main.ttw saved can lead to the GUID changing and the client thinking you connect to a new world.

And finally an error that maybe be a glitch but dont know it appears that in some point after a certain days the server is like it stucks in a single day, in the last map we were about day 360, suddenly one day a made a quick restart and the server went back to day 348, this happened everything I restarted the server.

World time is also part of the main.ttw. All this sounds like you got an issue with that one saving. A log might give more hints on what's going wrong.

 

[ScubaSteve]

Please tell me what I'm doing wrong :crushed: ...

I've setup the browser map before on many installations on my private server. I'm running Ravenhearst mod, and I have Alloc's fixes and ServerTools in place as well. I am behind a double-NAT only because CenturyLink's gateway is crap and can't handle my network of devices, so I have my router behind their gateway. I have several other services running through the double-NAT and they works just fine; I just forward the needed port in their gateway to my router's IP, then in my router I point it to my server's internal IP.

I can browse to http://<server's internal IP>:8082/static/index.html and pull up the browser map fine. Everything works great! I have most of the permissions set to '2000' level so everyone can see most of the map information without having to login to Steam. When I go to http://<my public IP>:8082 it doesn't pull up. I don't think it's a port forwarding issue. I have 26900 - 26902 UDP/TCP open for the server to appear on the server browser, and 8082 for the browser map. The Command Console it also set to 8080, so I know 8082 is the right port to forward. I even went to canyouseeme.org and it says it sees my service on port 8082.

I'm stumped!!

This is what I get when I try the public address in Chrome. I think that's just a generic "can't find the page" message.

================

This site can’t be reached

<my public IP> refused to connect.

Search Google for <my public IP> 8082 static index

ERR_CONNECTION_REFUSED

 

[Guppycur]

Make sure Windows firewall and/or other software firewalls are set up as well.

 

[ScubaSteve]

Would that really make a difference between internal and external access, if it works from a local IP?

I just tried it, rebooted too... no dice.

 

[Prisma501]

Make sure Windows firewall and/or other software firewalls are set up as well.

^ This. When first setting up a dev server on an old desktop and portforwarding the crap out of it all with no succes, it was as simple as windows firewall blocking the traffic. Gave me headaches back then.

Cheers

 

[Prisma501]

might wanna try 1(or all) of the fixes listed on here: https://www.quora.com/How-do-I-fix-ERR_CONNECTION_REFUSED-on-Chrome

Hope 1 of m helps.

Cheers

 

[ScubaSteve]

Thanks, but I don't think any of these are the solution.

 

[Prisma501]

Thanks, but I don't think any of these are the solution.

Okey by me. Im not the one being stumped

Cheers

 

[ScubaSteve]

Okey by me. Im not the one being stumped
Cheers

I'm not trying to be ungrateful, but it was kind of a "let me google that for you" assist. I looked at the site. Most are related to web hosting provider issues, but I'm not running Apachy or any web hosting services, and it's not a DNS issue. I also checked my hosts file and it's clean.

 

[StompyNZ]

How are you doing the double nat? Sounds like something isnt configured correctly there.

have you checked the router logs or done any traces?

what is your lan router model?

Are you on an external network when you try to view the map via external ip? most routers dont handle doublebacks on the same ip well when doing nat (you are coming from the wrong direction for the nat to work, because it maps from the external port to the internal)

 

[Lonestarcanuck]

Could it be as simple as not having your home machine access setup through your DMZ? I had that before.

when you say you can see http://serveriport but not publicIPort I found that was the issue.

maybe doing a trace route from public to see where it times out.

worse case call my wife, she is a veteran network administrator who wrote the book on TCP/IP (search for Kelli Adam and TCP/IP on Amazon) but she will charge you $100s in starbucks gift cards for 5 minutes of work

 

[ScubaSteve]

How are you doing the double nat? Sounds like something isnt configured correctly there.
have you checked the router logs or done any traces?

what is your lan router model?

Are you on an external network when you try to view the map via external ip? most routers dont handle doublebacks on the same ip well when doing nat (you are coming from the wrong direction for the nat to work, because it maps from the external port to the internal)

I have a piss-poor CenturyLink Technicolor C2100T as the main gateway that's only handling two IPTV boxes, then an ASUS RT-AC66W as the main router handling all the wireless and wired connections for the rest of the network. Previously, I had bridged the C2100T and used my Asus as the main router, but then the IPTVs stopped working after a while. CenturyLink came out and replaced the gateway but said it was going to happen again eventually. I still tried to bridge the new gateway and the ♥♥♥♥ty hardware would not allow it to happen so I left it alone and went with double-NAT :'(

When I access the browser map internally (when it works) both my gaming rig and the server are hooked up to the Asus router getting 192.168.1.x IPs. I have them on the same Trendnet 5-port switch in the same room. I've also tried power-cycling both the gateway and the router a few times to make sure they didn't need a reboot the put the changes into effect. I know that for the C2100T even thought you apply the changes it requires a reboot for them to go into effect.

I have several other services on the same server being hosted and accessible externally just fine: Plex, FTP, 7D2D, and more. My friends can login to 7D2D just fine and we have put in many hours, we just can't get to the browser map (externally).

- - - Updated - - -

Could it be as simple as not having your home machine access setup through your DMZ? I had that before.
when you say you can see http://serveriport but not publicIPort I found that was the issue.

maybe doing a trace route from public to see where it times out.

worse case call my wife, she is a veteran network administrator who wrote the book on TCP/IP (search for Kelli Adam and TCP/IP on Amazon) but she will charge you $100s in starbucks gift cards for 5 minutes of work

Sorry, but I am absolutely not going to setup a DMZ to my server. That's a bad security strategy. I sure hope your wife doesn't find out you're using DMZ :tickled_pink: Not that I have anything to hide, but there are plenty of malicious buffoons out there looking to just make a mess of peoples' lives.

My guess is this has something to do with my double-NAT and limitations of the web browser, but I don't understand why and wanted to make sure I've done everything I could.

 

[sniper670]

As long as you use the exe and not the patched Assembly-CSharp it should be fine. If you want to use DerPopo's Assembly-CSharp you have to apply my patches manually.

plase help me how i put them in manually i dont know how plase help me

 

[insG_Drifter]

Developed an issue recently. I did not notice until last night when I put up a clean server that when attempting to view the web map it’s fine until I try to log in.

After making the attempt to log in through Steam it simply returns “login failed”. Now the one server is newly installed but the other has been working along fine. I am admin on both servers so that shouldn’t be the issue. What file to I need to check to see what’s going on? Is this on steam only where the issue lies? What can I post to help determine the issue?

http://23.95.82.66:8282/static/index.html Is one of the ip’s. Does this seem likely to be a port issue maybe?

 

[xyth]

Developed an issue recently. I did not notice until last night when I put up a clean server that when attempting to view the web map it’s fine until I try to log in.
After making the attempt to log in through Steam it simply returns “login failed”. Now the one server is newly installed but the other has been working along fine. I am admin on both servers so that shouldn’t be the issue. What file to I need to check to see what’s going on? Is this on steam only where the issue lies? What can I post to help determine the issue?

http://23.95.82.66:8282/static/index.html Is one of the ip’s. Does this seem likely to be a port issue maybe?

I'm getting the same error last night, continues today...maybe a change on Steams end, as nothing changed on my server side.

 

[insG_Drifter]

I'm getting the same error last night, continues today...maybe a change on Steams end, as nothing changed on my server side.

Sounds like it if there’s more of us then. Thank you for the response. Wonder if it’s an api change again but guess we wait and see.