03-23-2011
TCP port numbers reused
Hello all,
I am loosing the diameter connection between two servers and when the connection is trying to comes up again i see the following message in the tracer.
[TCP port numbers reused] and after 1 second the connection resets.
As far i know the connection stay in state TIME_WAIT for a while(60 seconds in my case) and during this time the slot of the port is still reserved and no new connections is allowed to this port.
I tried to tune the TCP stack and change the time of TIME_WAIT state to 10 seconds using
ndd -set /dev/tcp tcp_time_wait_interval 10000
But the problem still exists.Even after 2 minutes if a connection comes up at the same own port then i receive TCP port numbers reused and then Reset.
Any idea?Maybe any other parameter that can modify?
Thanks in advance
10 More Discussions You Might Find Interesting
1. IP Networking
I am trying to connect via DBACCESS and Informix server to a server on a different computer. When I execute the connect command from dbaccess I get the following message,
Exec format error cannot bind a name to the port.
As far as I know the port is not being used by another client.
How... (1 Reply)
Discussion started by: lopez
1 Replies
2. UNIX for Dummies Questions & Answers
Hi All, I suspect this is simple but I cannot find any info on it.
I have a logfile on a solaris box (EMS) that I want to tail -f but I want the output of this to be redirected to a TCP port.
I have a second solaris box (PEM) running patrol enterprise manager that I am using as an alarm... (1 Reply)
Discussion started by: mscomms
1 Replies
3. UNIX for Dummies Questions & Answers
Hi
Is there any way to restrict the TCP-IP port usage.
I want to restrict TCP-IP port 1500/1550 to the oracle osuser.
Tanks in advance.
Remi (2 Replies)
Discussion started by: remivisser
2 Replies
4. SCO
sco unix 5.0.x, several weeks ago, I add a telnet-like service in inetd.conf, it runs well for 100 network terminals.
But nowdays, the terminals can connect to server successfully after booting machine, but several hours later, can not connect. "netstat -p tcp -a" can NOT find the port in... (1 Reply)
Discussion started by: shark_gao
1 Replies
5. Solaris
Hi all
I haven't had much experience with Solaris 10 but we've started running into a problem where a process hangs, it is killed and leaves a zombie process. The sysadmins are saying this zombie process is locking a tcp port and not allowing the process to start up.
The process is a usually... (5 Replies)
Discussion started by: casphar
5 Replies
6. UNIX for Dummies Questions & Answers
Hello, I have a service running (ODBC) and every now and then it will hang and I will have to stop and restart the service. The problem is when I stop the service, it indeed stops the service, but netstat reports a tcp port still open with the fin_wait_2 status. Then I must close the client... (1 Reply)
Discussion started by: raidzero
1 Replies
7. Shell Programming and Scripting
I have multiple processes running the same program on my linux machine. For each process I want to be able to use a unique (available) TCP port. I have thought of using netstat to check which ports are available for use however, the time-window between checking and selecting might expose some race... (1 Reply)
Discussion started by: timmylita
1 Replies
8. Programming
Does anyone know if there is a C API call to get the status of a TCP port? As opposed to running netstat and parsing the results. At the moment I have to attempt to bind() and pick up on the address in use error which isn't very elegant
Thanks
---------- Post updated at 10:42 AM ----------... (0 Replies)
Discussion started by: janra
0 Replies
9. Programming
Hello,
I'm trying to write a small c application to test a tcp port. This works fine for the most part but the default timeout on the connect is very long. I have been reading many posts but and it looks like I need to set the socket to be non-blocking and poll for a result. I have been totally... (2 Replies)
Discussion started by: tjones1105
2 Replies
10. IP Networking
i want to kill a tcp connection by killing its pid
with netstat -an i got the tcp ip connection on port 5914
but when i type ps -a or ps-e there is not such process running on port 5914
is it possible that because i do not log on with proper user account i can not see that process running? (30 Replies)
Discussion started by: alinamadchian
30 Replies
LEARN ABOUT LINUX
tcp-environ
tcp-environ(5) File Formats Manual tcp-environ(5)
NAME
tcp-environ - TCP-related environment variables
DESCRIPTION
The following environment variables describe a TCP connection. They are set up by tcp-env, tcpclient, and tcpserver. Note that TCPLOCAL-
HOST, TCPREMOTEHOST, and TCPREMOTEINFO can contain arbitrary characters.
PROTO
The string TCP.
TCPLOCALHOST
The domain name of the local host, with uppercase letters converted to lowercase. If there is no currently available domain name for
the local IP address, TCPLOCALHOST is not set.
TCPLOCALIP
The IP address of the local host, in dotted-decimal form.
TCPLOCALPORT
The local TCP port number, in decimal.
TCPREMOTEHOST
The domain name of the remote host, with uppercase letters converted to lowercase. If there is no currently available domain name for
the remote IP address, TCPREMOTEHOST is not set.
TCPREMOTEINFO
A connection-specific string, perhaps a username, supplied by the remote host via 931/1413/IDENT/TAP. If the remote host did not sup-
ply connection information, TCPREMOTEINFO is not set.
TCPREMOTEIP
The IP address of the remote host.
TCPREMOTEPORT
The remote TCP port number.
SEE ALSO
tcpclient(1), tcpserver(1), tcp-env(1), tcp(4)
tcp-environ(5)