Parallels

Product: Ensim Pro 4.0.3 and above
Date: 14-Jun-2007

Issue:

When the cron file /etc/cron.daily/1hup_apache_logs is executed the httpd service is not able to reload properly and it get stuck. The service is dead and it needs to be restarted manually. Even the server load goes high. And in some cases the server even crashes.

The error reported is:
============
[Wed Jan 17 18:00:13 2007] [warn] child process 11251 still did not exit, sending a SIGTERM
[Wed Jan 17 18:00:30 2007] [warn] child process 22053 still did not exit, sending a SIGTERM
[Wed Jan 17 18:00:30 2007] [warn] child process 1069 still did not exit, sending a SIGTERM
=============

Resolution:

Change "reload" to "restart" in /etc/cron.daily/1hup_apache_logs

Content of /etc/cron.daily/1hup_apache_logs are as follows:

#!/bin/bash
#/etc/rc.d/init.d/httpd_app_init reload-grc
# XXX Fix of the "lost child came home" issue
! /sbin/chkconfig httpd || /etc/rc.d/init.d/httpd_app_init reload

Change to

#!/bin/bash
#/etc/rc.d/init.d/httpd_app_init reload-grc
# XXX Fix of the "lost child came home" issue
! /sbin/chkconfig httpd || /etc/rc.d/init.d/httpd_app_init restart