Applicable versions and levels
Windows versions: V5.0L10 to V10.0L21
Solaris versions: 5.0 to 10.1
HP-UX versions: 10.0
Action
Points to check
Was the holiday calendar changed in a state where all subsystems were started?
Cause
It may have been determined that the subsystem environment does not exist.
If there is a subsystem structure in the cluster operation, and that subsystem is stopped, the subsystem environment that is placed on a shared disk, such as the Systemwalker Operation Manager database, will have unmounted status.
The changes made to the holiday calendar are reflected in the subsystem environment.
If a subsystem has not started, it will be determined that the environment for that subsystem does not exist, so the changes made to the holiday calendar are not reflected in the startup day of the job net.
Action method
Start all subsystems, and then modify the calendar in the same way on both the active node and the standby node. The procedure is explained below.
Check the subsystem number that has not started.
Check the Systemwalker Operation Manager running status using the process running status display command (mppviewo):
[Windows versions]
<systemwalker install directory>\mpwalker.jm\bin\mppviewo
[UNIX versions]
/opt/systemwalker/bin/mppviewo
Start the subsystems that have not started.
Follow the example below to perform the operation when starting the subsystem for any subsystem number manually.
[Example] Starting subsystem n
[Windows versions]
Start the following Systemwalker Operation Manager services from the Services dialog box opened from Control Panel:
- Systemwalker MpMjesn
- Systemwalker MpJobschn
Note:
Systemwalker MpMjesn and Systemwalker MpJobschn must be started continuously.
Start the services with the same subsystem number as MpMjes and MpJobsch.
[UNIX versions]
/bin/sh /etc/opt/FJSVMJS/etc/rc3.d/S99MJS -sys n |
Note:
S99MJS and S99JOBSCH must be executed continuously.
In the -sys parameter, specify the same subsystem number as S99MJS and S99JOBSCH.
If the subsystem cannot be started, restart Systemwalker Operation Manager or restart the system.
Change the calendar on all cluster systems.
