Hi,
I have a perl script (on AIX433) from which I first fork out a process and then consequently setsid() it. After forking, the main perl script keeps running, checking the log written to by the forked out process, and taking actions based on it. The forked out process is a server process that is supposed to listen on a specified port, and act on client requests on that port.
Now, at times, I have a situation where the terminal where the script is running just hangs, and so does the forked out process. No amount of interrupts or kill -9 from other terminals bring this terminal back to life, or get the processes responding again. More surprisingly, a netstat check on the port where the *.LISTEN from the server process is to be listening doesn't seem to be showing any process occupying this port, eventhough a ps check shows the server process to be running! This behaviour recurs entirely at random!
Any clues on what is going on here? Could this be happening because of some system limitation on the AIX box, such as load on the machine, max number of open filehandles etc? What is the equivalent of sysinfo for AIX, if I wanted to check? Could it be due to the telnet client used to create the terminals and connect to the server?
Thanks in advance.
regards,
BB.
I have a perl script (on AIX433) from which I first fork out a process and then consequently setsid() it. After forking, the main perl script keeps running, checking the log written to by the forked out process, and taking actions based on it. The forked out process is a server process that is supposed to listen on a specified port, and act on client requests on that port.
Now, at times, I have a situation where the terminal where the script is running just hangs, and so does the forked out process. No amount of interrupts or kill -9 from other terminals bring this terminal back to life, or get the processes responding again. More surprisingly, a netstat check on the port where the *.LISTEN from the server process is to be listening doesn't seem to be showing any process occupying this port, eventhough a ps check shows the server process to be running! This behaviour recurs entirely at random!
Any clues on what is going on here? Could this be happening because of some system limitation on the AIX box, such as load on the machine, max number of open filehandles etc? What is the equivalent of sysinfo for AIX, if I wanted to check? Could it be due to the telnet client used to create the terminals and connect to the server?
Thanks in advance.
regards,
BB.