Apache Service failed to respond within 60 seconds.

  • 7012737
  • 27-Jun-2013
  • 28-Jun-2013

Environment

NetIQ Access Manager 3.2.1-57 + IR1-201
MAG Service running on SLES 11 SP2.

Situation

Linux Access Gateway Service (AGS) running on SLES 11 SP2.
In order to mitigate a PCI compliance vulnerability found in ActiveMQ listening on port 61617, KB 7010673 was followed.  Basically, ActiveMQ was modified to listen only on the loop back.  After restarting ActiveMQ on each MAG, the admin console now shows the following:

Apache Service failed to respond within 60 seconds. 
(Required Action) Check log files, and verify all services are running.

The catalina.out file on the Access Gateway Service showed a number of RequestTimeout() messages at the time.  

Restarting ActiveMQ again, made no difference.  Backing out changes and restarting ActiveMQ, made no difference.  Restarting the server completely made no difference.

While this shows as an error, all services on the AGS's seem to be functioning correctly even with this error in the admin console.

Resolution

Manually restart /etc/init.d/novell-apache2 after the server reboot.

ActiveMQ could not contact the Apache proxy  at load time for some reason. By manually loading the Apache proxy after the server reboot, ActiveMQ was able to contact the Apache server and initialise correctly.

By adding a new script to auto restart Apache in the rc3.d level (using chkconfig or insserv), the restart of apache can be automised.

Cause

There may be some issue with the load order and/or timing of ActiveMQ and Apache loading.  Apache may need more of a delay when starting.