"Error, Either an old instance of teamspeak is still running..."

[Degas]

Need some help with this, please!

For some reason, my TS Linux Server (newest version), would not allow any users to connect this morning. We have a standby server that we use, and we can all log in there just fine.

I could log into admin/superadmin with no problems. This server is on a shared host, by the way. In superadmin, the servers showed up just fine, and I could start it and stop it. I could also start and stop it normally from the SSH shell access (I use Putty). However, I could not login, even using telnet commands. It just kept giving me an "Error, no server specified" message. When I did a server list command, the server (8855) showed up just fine. But when I did the "login 8855" it kept giving me the "no server specified" message.

OK, so it's a fairly new installation, with only a few users and admins. I figured I'd just do a re-install. When I tried to do that, I got an error message about the name of the .tar file, and it kept renaming it to .tar.out, or some such.

I finally figured I'd just bite the bullet and do a fresh install. So, I deleted the entire "tss2_rc2" directory from within the "tss2" folder I had in my root directory. I then extracted the files again. When I went back into Putty and tried to start the server, I got the following error message:

Error, Either an old instance of teamspeak is still running, or an other application is using the tcpquery port!
Error, Server was not started!

So, I tried to do a server STOP, and got the following message:

tsserver2.pid is missing, no started server ?

I'm scratching my head by this time, and decided to see what would happen if I tried to log into the (now non-existent) TS Server on port 8855. Sure enough, using the old superadmin password, I was able to log in. And it showed the server on port 8855 still running! I stopped it, and tried to start a server from SSH in the fresh TS Server tss2_rc2/ folder. Got the same messages as above. So, I figured maybe I would just try a whole new directory. I transferred the extracted .tar file into a new folder, with a new name (tss). I completely deleted the old (tss2) directory where I had originally installed the server software. I then went through the install process in that new (tss) directory. This installed a fresh tss2_rc2 folder in the fresh directory. However, when I head to that directory via SSH and try to start the server, I get the same error messages listed above.

I've been searching the TS forums all day for this error message, and really have seen anything to help. One thing I DID try was to run

netstat -ap

Which showed nothing at all using any ports, more specifically nothing was using port 8767.

I'm tearing my hair out here. If it was my server, I'd try rebooting it and seeing if that got rid of the "old instance" that is supposedly running. Since it's not my server, I don't have a clue what to do next.

Can anyone help?

[Degas]

linuxfan, save me! LOLOL

I checked for running processes in mysql (through the server admin panel), and it didn't show a thing.

I saw a linux line command somewhere that will do the same thing, but I can't find one that will work. I tried

ps -af

and get the message that there is no such command as "ps", which probably means it's not enabled for admins on a shared linux server. I'm wondering if my next move isn't going to be to beg the super admin to reboot the server.

[Degas]

This is a bad problem for us, and we need some support from the developers.

[Brain]

Degas, try "help login", you seem to have a strange idea about the TCPQuery command syntax.

And if you want to terminate all running TS2 servers promptly try "killall server_linux".

[Degas]

LOL, Brain. Thanks for the reply.

I don't have a "strange" idea about TCP query command syntax...I have NO idea :D Well, I'm learning.

I've tried killall server_linux and kill -9, etc. Still get the "tsserver2.pid is missing, no started server ?" message. I think it's a resident memory error. Since I removed the TS folder with a virtual server still running, Linux thinks there is still an "old instance" up.

We are 90% sure that rebooting the hardware will fix this. Now it's just a matter of convincing the host administrators to do it.

Just for info, here is the terminal window info as it appears when I try to start the server:

login as: username
username@www.mydomain.com's password:
Last login: Sun Sep 26 20:17:34 2004 from *************
-bash-2.05b$ ls
mainwebsite_cgi mainwebsite_html tss2
-bash-2.05b$ cd tss2
-bash-2.05b$ ls
ts2_server_rc2_20201.tar tss2_rc2
-bash-2.05b$ cd tss2_rc2/
-bash-2.05b$ ls
INSTALL changelog.txt mysql_sql sqlite_sql
INSTALL.mysql httpdocs server.ini tcpquerydocs
Manual libsqlmy.so server_linux teamspeak2-server_startscript
README manual.html sqlite.so
-bash-2.05b$ ./teamspeak2-server_startscript start
starting the teamspeak2 server
Error, Either an old instance of teamspeak is still running, or
an other application is using the tcpquery port!
Error, Server was not started!
-bash-2.05b$ ./teamspeak2-server_startscript stop
tsserver2.pid is missing, no started server ?
-bash-2.05b$

[Brain]

I think I see your problem. Commands like killall or the famous kill -9 only show their true power when you have root privileges :)

[Degas]

yeah, exactly...which is why I've been exchanging emails with the host admins for a week trying to get them to reboot the server, LOL.

When we get the inevitable server restart, I'll update the thread on whether or not it fixed our problem.

Oh, thanks for the "help login" tip. Using telnet, I got logged in as superadmin to TS last night, and confirmed that there is no TS server running, and no jobs or pids active at all.

Our Linux guru has been unable to help with any of this, due to workload IRL. Finally got with him last night and he's certain a reboot will correct the problem.

UPDATE: Restarting the hardware solved the problem of the "old instance". Just a tip...NEVER remove the Teamspeak folder without stopping the TS server :P

[PoMM3]

I have the same problem, I don't want to reboot... Any idea?

Thank you for your responses.

[ICEPOC]

i have this same problem.

except i am root owner.

i have rebooted same box.

no TS running at all but wont let me boot up TS and says:

Error, Either an old instance of teamspeak is still running, or an other application is using the tcpquery port! Error, Server was not started!

any idea's ?

[Dimedo]

It seems I have the same problem here. I'm running (or trying to) TeamSpeak 2 Server (2.0.23.15 BETA) on Debian GNU/Linux Etch (4.0 RC1) with MySQL (5.0.32) database backend. Until today it worked for weeks or even months without any problems. As far as I can say, there was nothing changed related to the TeamSpeak 2 Server filesystem or database structure. Also I can say, it was not stopped by hand, it simply wasn't running when I came home from work. The last log entries look like someone had stopped the server intentionally:

02-03-07 14:25:34,ALL,Info,server, Server shutdown initialized
02-03-07 14:25:34,ALL,Info,server, Stopping VirtualServer id:3 with port:8768
02-03-07 14:25:34,ALL,Info,server, Stopped VirtualServer id:3 with port:8768
02-03-07 14:25:34,ALL,Info,server, Stopping VirtualServer id:1 with port:8767
02-03-07 14:25:34,ALL,Info,server, Stopped VirtualServer id:1 with port:8767

The error message is also displayed by the server binary without any arguments. So there seems to be no problem with the start script. Here again, the message:

Error, Either an old instance of teamspeak is still running, or
an other application is using the tcpquery port!
Error, Server was not started!

The server.ini says TCP Query port is 51234, and it IS unused currently. So it has to be something else. There is no instance of server_linux running. I also tried to start the old 2.0.20.1 binary, but it displays the same error as mentioned above. Changing the TCP Query port in the server.ini to something else has not changed anything on the problem.

This (http://194.97.1.54/showthread.php?p=135195) thread is also about the same error. There are some more threads available by Google search. But there is not one solution to this problem. Not one of the threads even mentioned if the hardware reboot worked. A hardware reboot isn't a nice thing on a remote linux box which has lots of other services running, but if I don't find an other idea, there will be nothing else to do for me.

If you have any idea, please post them.

[Dimedo]

I solved the problem through a full hardware reboot.

An ugly and rough way, indeed. That no TeamSpeak developer has said anything on this problem since 2004 is evidence again that TeamSpeak is badly coded and isn't supported very good. The only help I have ever found in this forum came from other frustrated users. I can only hope that TeamSpeak 3 will be much better in quality and support.

[tytyrus]

I solved the problem through a full hardware reboot.


I had the very same problem. Maybe saying that I got the same error message is better.

I found out that (oddly enough) the problem was a full disk partition. The server couldn't open the ini file for read/write:

open("/home/tss2_rc2/server.ini", O_RDWR|O_CREAT|O_TRUNC, 0666) = 6

write(6, "[Main Config]\nBoundToIp1=\n\n", 27) = -1 ENOSPC (No space left on device)

If you have enough skills to understand its output, I suggest using strace. that's how I got to the answer.

strace ./server_linux -PID=tsserver2.pid

I don't subscribe to the bad comment about teamspeak. my only gripe about it (although it's no small thing IMO) is that it's not Free Software.

I will concede that more sensible error condition detection and messages would help troubleshooting, though. :D

-andrew (http://andrew.org)

[Peter]

Nice find tytyrus!

Though the error is probably not worth fixing at this stage (development of ts3 is more important) it certainly sheds some light on the cause of this error. If we need to make a new TS2 server binary (due to some critical problem found which needs attending to) we will consider fixing this if it can be done quickly (which I suspect, without knowing the code).

[biovore]

I have had this same problem. But I have noticed that this error is a generic error for the TCP socket failling to bind to a interface. In my case I had serveral aliased interfaces that wern't up do to a reboot, and when I tried to bring TS up it gave me the error in question. After I setup the interfaces and gave it the correct IP to bind to, it worked fine here..

[Reedy Boy]

Im currently fiddling about trying to get this to work on an NSLU2

As the device is ARM, its running through Qemu (We have successfully got vent to work, we just want TS to too!)

I get the
Error, Either an old instance of teamspeak is still running, or
an other application is using the tcpquery port!
Error, Server was not started!


But, i dont get this till i press ctrl + c to try and stop the process

I know theres nothing using the ports, ive tried binding it to 0.0.0.0 - ie all interfaces and its ip - 192.168.0.71 and i've tried rebooting


Anyone any ideas?


I did manually create the server.ini from a copy of my windows one to see if that would help... But no avail

Its not getting as far as creating the server.log either....

[Ciberputa]

Hi

I have the same problem. I'm behind a Router into a LAN, so, I open the 8767 UDP port, 14535 for web admin and 51234 for TCP/IP query. Really, only 8767 is necessary.
When I start the TS, everything seems ok, but nobody outside the lan is able to connect.
So, I edit the server.ini and put my external IP to "bind to ip". Stop the server and start, but then:

Error, Either an old instance of teamspeak is still running, or an other application is using the tcpquery port.


There is not any TS running and there is not any application using the port 51234. IF I change again the "bound to ip" to the internal IP (192.168.1.1) or leave It blank, TS starts again.

Here is my server.ini:

[Main Config]
BoundToIp1=212.60.17.199
ExternalIPDectection=1
HTTPServer Port=14535
HTTPServer Enabled=1
DateTimeFormat=dd-mm-yyyy hh:nn:ss
TCPQueryPort=51234
AllowedClientNameChars=
DisAllowedClientNameChars=()[]{}

[debug]
MessageTypes=LMTALL
MessageDepths=LMDALL

[WebPost]
AdminEmail=digitalia@mundor.com
ISPLinkURL=http://grupo97.halconesrojos.com
ISPName=Grupo 97
ISPCountryNumber=1121
Enabled=1
PostURL=1
ListPublic=1
UserAgent=teamspeak

[log]
access_r=0
access_u=0
channel_registerred=0
channel_unregisterred=0
sa=0
chat=0
kick_server=0
kick_channel=0

[Spam]
max_commands=10
in_seconds=2


Any help?

[Jacker]

I received this same error and was able to kill the process and then restart the server. I used the command `kill 25278` where 25279 was the PID of the process holding the port open or the teamspeak server. So I did not have to reboot the hardware to get this working.

Hope that helps someone!

[StaticSt0rM]

Check if your IP is OK.

It happened to me when I changed the machine IP in the interface BUT not in the TS configuration.

If the IP is wrong, the erros is the same as if the PORT is already occupied.

;-)