Red Hat / CentOS TIBCO Spotfire server will not start after a reboot

Red Hat / CentOS TIBCO Spotfire server will not start after a reboot

When a TIBCO Spotfire Server running on Red Hat Linux or CentOS encounters an abrupt shutdown, the service may not resume after a reboot. This happens under circumstances where the Java process for TIBCO Spotfire Server is killed without undergoing the usual shutdown routine.

This issue is caused by a PID file which is located in /opt/tibco/tss/10.3.x/tomcat/bin/catalina.pid being left behind after the shutdown. This file is normally removed by the systemctl shutdown command (which runs the tomcat shutdown.sh)

The startup command, systemctl start tss-10.3.x checks for the presence of old PID files and will then abort startup if one is found. It does this without checking if the PID actually relates to a running process or not.

This is a bug which has been fixed starting from TIBCO Spotfire server versions 11.4 or higher. However for lower versions the following workaround can be implemented.

1. Edit the file /etc/init.d/tss-10.3.11 (Take a backup first)
2. Comment out the line rh_status_q in the start option

No further action is needed and the next time an abrupt shutdown is encountered, the PID will automatically be cleared and the service will startup as usual.