Top
Systemwalker Operation Manager Troubleshooting Guide
FUJITSU Software

4.14.23 A Jobscheduler Error Message is Displayed during Startup of the Cluster System [UNIX Versions]

Error message

MpJobsch: ERROR: 0027: The jobscheduler failed to retrieve day change time.

Applicable versions and levels

Action 1

Points to check

Was the symbolic link of the script that starts the calendar daemon set correctly?

The following method can be used to check if the symbolic link was set correctly:

  1. Move to the directory that creates the symbolic link

    #cd /etc/rc3.d
  2. Execute the ls command to check the S79JMCAL file properties

    If the "S79JMCAL -> /opt/FJSVjmcal/bin/S99JMCAL" part is displayed in the same way, it means that the symbolic link has been set.

    #ls -l
    lrwxrwxrwx    1 XXXXX  XXXXX   XX  mm:dd hh:mm S79JMCAL -> /opt/FJSVjmcal/bin/S99JMCAL

Cause

The Jobscheduler daemon started when the calendar daemon is not running or while the startup is in progress.

For the Jobscheduler daemon to start, the calendar daemon must already have been started.

If the environment is not a cluster environment, the calendar daemon and the Jobscheduler daemon will be started by the Systemwalker Operation Manager startup script when the system starts up, and the startup order is guaranteed. However, in a cluster environment, the Jobscheduler daemon is started from the cluster daemon, not from the Operation Manager startup script.

For this reason, the cluster daemon may start before Systemwalker Operation Manager, and the Jobscheduler daemon may start before the calendar daemon startup completes.

Action method

In cluster system, configure the settings shown below so that the calendar daemon starts before the Jobscheduler daemon that is started together with the cluster system startup.

On the active and standby nodes, execute the following command so that the calendar daemon startup script is called before the cluster daemon:

# cd /etc/rc3.d
# ln -s /opt/FJSVjmcal/bin/S99JMCAL S79JMCAL

Note that the above numeric value "79" represents the startup order of each daemon started in /etc/rc3.d. The daemon with the lowest numeric value is started first. The supported cluster daemons use 80 or more, therefore 79 is used. Refer to the operating system manual for information on how to set rc3.d.

Action 2

Points to check

Has the calendar daemon (f3crhcs2) stopped?

Check whether the process is running:

# ps -ef |grep f3crhcs2

Action method

If Systemwalker Operation Manager was stopped using the poperationmgr command, execute the soperationmgr command to start Systemwalker Operation Manager (calendar daemon), and then with that status start the Jobscheduler daemon from the cluster system.