While running some kind of automated tests on our AIX 6.1 machine we are facing the problem that a port seems to be in use although the corresponding process has already terminated.
E.g. if we kill a hanging application using port 9999 we’ll get the following:
user@aix:~/EP$ netstat -Aan | grep 9999
…
f1000700149ec3b0 tcp 0 0 *.9999 *.* LISTEN
…
user@aix:~/EP$ rmsock f1000700149ec3b0 tcpcb
Wait for exiting processes to be cleaned up before removing the socket
Currently, we only can recover from this state by rebooting the machine.
We already tried to set SO_REUSEADDR but we still get this strange kind of error.
The error is not application specific - it happens with different applications.
user@aix:/etc$ oslevel -s
6100-04-01-0944