0

We have an application currently crawling the web on a dedicated CentOS server, using Plesk. The application makes use of proc_open and exec, and it usually works but eventually fails. The error logs sometimes shows 2 interesting errors: Unable to fork followed by a sequence of pipe errors - other times nothing.

Like I mentioned before, the problem is that the application suddenly and randomly stops after crawling for a few hours/days, and usually there's not even proper debugging / profiling information in the PHP error logs.

Is there any application, software, suit, method, or any other kind of solution to monitor the server resources and errors, and get a full idea of what could be happening that makes the application stop?

dimo414
  • 47,227
  • 18
  • 148
  • 244
Chris Russo
  • 450
  • 1
  • 7
  • 21

1 Answers1

2

You may want to try NewRelic for monitoring your server and software.

Your forking issue may be a file descriptor limit or PID limit of the machine.

Try increasing the PID max

echo 5000000 > /proc/sys/kernel/pid_max

drmarvelous
  • 1,663
  • 10
  • 18
  • Thanks for the feedback, but I already tried both my friend, there's no error information at all at this point. NewRelic doesn't provide any information regarding the interruption. As well, I tried to keep track of the memory usage, it doesn't reach any limit... I'm lost here. – Chris Russo Jul 31 '14 at 02:21