Top
Systemwalker Operation Manager  User's Guide
FUJITSU Software

1.2.3 Shutting Down the Systemwalker Operation Manager Servers [UNIX version]

When you carry out the shutdown procedures, all of the daemons for the various Systemwalker Operation Manager functions are automatically stopped and the Systemwalker Operation Manager server terminates its operations.


How to shut down manually

If you need to manually shut down a Systemwalker Operation Manager server for some reasons such as maintenance work, daemons of each function can be shut down in a batch with the poperationmgr (daemon termination) command. For details on the poperationmgr command, see the "Shared Commands" section of the Systemwalker Operation Manager Reference Guide.

Note

Coexistence with Systemwalker Centric Manager

To terminate the Systemwalker Operation Manager in an environment with Systemwalker Centric Manager is used together, terminate all the Systemwalker Centric Manager daemons as well. For the daemon termination commands for Systemwalker Centric Manager, see the Systemwalker Centric Manager Reference Guide.

Note

Stopping the Jobscheduler daemon

Note the following when stopping the Jobscheduler daemon.

  • If a job net having the Interstage attribute is running

    Do not stop the Jobscheduler daemon independently. Cancel the job net before stopping the Jobscheduler daemon.

  • When stopping the process (tskmnsrv)

    Do not use the "kill -9" command when stopping the process (tskmnsrv). Otherwise, the Jobscheduler daemon cannot be stopped. Additionally, the Jobscheduler daemon may not work correctly in a cluster system.

    If the process (tskmnsrv) is inadvertently terminated using the kill -9 command, restart the system.

Note

Stopping the daemons at shutdown in Red Hat Enterprise Linux 7

In Red Hat Enterprise Linux 7, the operating system may need to be stopped or restarted in the cases below even if the daemons are configured to stop automatically when the operating system is stopped. In these cases, manually execute the poperationmgr service/daemon shutdown command in advance to stop the Systemwalker Operation Manager daemons.

  • The daemons were manually restarted (using the poperationmgr/soperationmgr command, for example).

  • The daemons were restarted using daemon control in the environment setup window.

  • The backup command was executed.

  • A policy was distributed with "Apply once" specified.

In Red Hat Enterprise Linux 7, when the operating system is shut down or restarted without stopping Systemwalker Operation Manager, the behavior will be as explained below. This is identical to the behavior when the system fails during operations.

  • The startup mode for the Job Execution Control function at the next startup will be recovery mode.

  • Waiting demand jobs will be re-executed at the next startup.

    Processing of executing demand jobs depends on whether Re-execute Job is specified.

    If Re-execute Job is specified: Jobs will be re-executed from the beginning at the next startup.

    If Re-execute Job is not specified: Jobs will not be re-executed.

  • Scheduled jobs will not be re-executed at the next startup.

  • If the execution server for network jobs or distributed execution jobs was shut down, neither demand jobs nor scheduled jobs will be re-executed, regardless of the status of the jobs at shutdown and whether Re-execute Job is specified.

Enterprise EditionHow to shut down multiple subsystems manually

With Systemwalker Operation Manager Enterprise Edition version, multiple subsystems can be started up. To manually shut down subsystems with corresponding subsystem numbers, follow the example below.

Example 1: To shut down subsystem 1

/bin/sh /opt/FJSVJOBSC/etc/rc0.d/K00JOBSCH -sys 1
/bin/sh /etc/opt/FJSVMJS/etc/rc0.d/K00MJS -sys 1

Note)

  • Make sure K00MJS is executed immediately after K00JOBSCH is executed.

  • For the -sys option of K00MJS and K00JOBSCH, specify the same subsystem number.


Example 2: To shut down all the subsystems

/bin/sh /opt/FJSVJOBSC/etc/rc0.d/K00JOBSCH -sys all
/bin/sh /etc/opt/FJSVMJS/etc/rc0.d/K00MJS -sys all


Information

K00JOBSCH and K00MJS storage paths

The path to the K00JOBSCH and K00MJS storage varies depending on the OS used as follows. Change those paths according to the OS you are using.

[K00JOBSCH]

Solaris

/opt/FJSVJOBSC/etc/rc0.d

HP-UX system:

/opt/FHPJOBSCH/bin

AIX

/usr/FAIXJOBSC/bin

Linux

/opt/FJSVJOBSC/etc/rc0.d

[K00MJS]

Solaris

/etc/opt/FJSVMJS/etc/rc0.d

HP-UX system:

/opt/FHPMJS/usr/lib/mjes

AIX

/opt/FAIXMJS/usr/lib/mjes

Linux

/etc/opt/FJSVMJS/etc/rc0.d