wurm dedicated server binding issues

[fireblade 5]

ive been trying to get a headless server on ubuntu server 14.4 LTS
 
i have tried configuting the wurmlogin.db ip addresses setup as 192.168.1.9 or 0.0.0.0

but the server insists on binding as ipv6 only

$ sudo netstat -a -p | grep LIST
tcp        0      0 *:ssh                   *:*                     LISTEN      9958/sshd
tcp6       0      0 [::]:7221               [::]:*                  LISTEN      4769/WurmServerLaun
tcp6       0      0 [::]:ssh                [::]:*                  LISTEN      9958/sshd
tcp6       0      0 gserver.corp.fire:48010 [::]:*                  LISTEN      4769/WurmServerLaun
tcp6       0      0 gserver.corp.fire:3724 [::]:*                  LISTEN      4769/WurmServerLaun

 

anyone have any suggestions?

 

ive also had issues with windows where the client appears to be trying to communicate via a vmware virtual network adapter for no obvious reason if i run the server locally rather than the interface withe the default gateway even when the defined server has an ip (192.168.1.39) in the same subnet as the default gateway

 

e.g it uses 192.168.74.1 (VMware Network Adapter VMnet1)

 

rather than

 

192.168.1.39 (Default Gateway: 192.168.1.100)

this might also be something to do with server binding as if i change its external ip address to 0.0.0.0

the client can connect

im just guessing but it appears that the server is not binding to the specifically defined ip address atleast in some cases




 

 

 

[LastChime 4]

Thank you, oddly enough that solved my connection refused error, set external ip to 0.0.0.0 . Hope your issue gets resolved.

[Turbine 1]

Hmm interesting, 0.0.0.0 usually binds to all addresses doesn't it? Strangely no matter which address I use, if any one uses port 27016 it fails to bind.

 

Disabling ipv6 doesn't help either (I believe it does ipv4 tunneling by default these days):

-Edit: /etc/sysctl.conf

Write:

net.ipv6.conf.all.disable_ipv6 = 1

net.ipv6.conf.default.disable_ipv6 = 1

net.ipv6.conf.lo.disable_ipv6 = 1

-Run: sysctl -p

 

Here's some useful things to help with debugging:

-Kill all threads when Wurm fails to close properly after a taken port error:

ps -ef | grep WurmServ | grep -v grep | awk '{print $2}' | xargs kill -9

-List ports listened on (27016 is important):

sudo netstat -ntulp | grep Wurm

Edited October 23, 2015 by Turbine

[fireblade 5]

just guesing it might be some thing to do with the server not binding on ipv6 and ipv4 simultaniously

if i remember right

 

0.0.0.0 bind to any ipv4 address

:: binds to any ipv6 address (and in java all addresses  see http://docs.oracle.com/javase/8/docs/technotes/guides/net/ipv6_guide/index.html#dual)

 

ill try binding to :: later and see what happens

 

i might be able to force ipv4 with java.net.preferIPv4Stack (-Djava.net.preferIPv4Stack=true) but will need to investigate that

 

i have a feeling that its not using the specified ip address directly but "processing" somehow first
 

Edited October 23, 2015 by fireblade

[erozaxx 37]

I found out, that not only server but also client has to have only ONE possible interface to connect to. In my case, if I disable VMnet interfaces and start the server, then enable VMnet interfaces and run client it does not connect. If I disalbe VMnet interfaces though, then run server and client it connects and then I can enable VMNet interfaces again - hope that helps.

[Nazaryx 0]

Im having the exact same issue at the moment, the wurm server running on my ubuntu instance is binding to the IPv6 ports and not listening to the IPv4 ports regardless of port forwarding.