Use this task to create application monitors and to link
them to one or more application controllers that might belong to a
resource group.
For the definition of application monitor, see the PowerHA® SystemMirror Glossary in
the AIX® Information Center.
Fields
- Application monitor type
- Specify one of the following application monitor types:
- Process
- The termination of one or more processes of an application is
detected by resource monitoring and control (RMC).
- Custom
- The health of an application is checked by a custom monitor method
at user-defined polling intervals.
- Processes to monitor
- Specify the unique names of one or more processes that are being
monitored. The output of the /usr/bin/ps -el command
provides the process name. This field can be set only for a Process monitor.
- Monitor script
- Specify the full path and file name of a method to check the application
status. This field can be set only for a Custom monitor.
You cannot specify arguments for this script.
- Process owner
- Specify the user name of the owner of the processes. This field
can be set only for a Process monitor.
- Instance count
- Specify the number of instances of a process to monitor. If more
than one process is specified in the Process to monitor field,
the instance count is 1. This field can be set only for a Process monitor.
- Monitor interval (seconds)
- Specify the polling interval in seconds for checking the health
of the application. If the application monitor does not respond within
this interval, the application is considered to be hung. This field
can be set only for a Custom monitor.
- Hung monitor signal
- Specify the signal, in numbers, that the system must send to stop
the Monitor Script if it does not return within
the time specified for the Monitor Interval.
This field can be set only for a Custom monitor
and the default value is 9 (SIGKILL).
- Stabilization interval (seconds)
- Specify the time interval, in seconds, for the PowerHA SystemMirror to wait
before performing the action. The stabilization period is used in
different ways based on the type of monitor mode that you select.
- When the Monitor mode field is set to Long-Running
Monitoring, the stabilization interval is the period during
which PowerHA SystemMirror waits
for the application to stabilize before it starts monitoring for the
successful running of the application.
For instance, with a database
application, you might want to delay monitoring until after the start
script and initial database search have been completed. Experiment
with this value to balance performance with reliability.
- When the Monitor mode field is set to Startup
Monitoring, the stabilization interval is the period within
which PowerHA SystemMirror monitors
that the application has successfully started.
When the time expires, PowerHA SystemMirror stops
monitoring the application startup and continues event processing.
If the application fails to start within the stabilization interval,
the resource group acquisition fails for the node and PowerHA SystemMirror begins
resource recovery actions to acquire the group on another node.
The
number of seconds you specify must be equal to the time it takes for
the application to start. This depends on the application you are
using.
- When the Monitor mode field is set to Both,
the application monitor waits for the specified length of time for
the application to start successfully. It uses the same interval to
wait before the application monitor begins to check that the application
is successfully running on the node.
- Restart count
- Specify the number of times to restart the application controller.
- Restart interval (seconds)
- Specify the number of seconds (in addition to the Stabilization
interval field) that the application must remain stable
before resetting the Restart count.
- Monitor mode
- Select one of the following modes for the application monitor:
- Action on application failure
- Select the action that the application monitor must take after
all restart attempts are exceeded:
- When you select Notify, the application
monitor runs the specified notify script and takes no further action.
- When you select Failover, the application
monitor moves the resource group that contains the monitored application
to another node in the cluster.
- Notify script
- Specify the full path and file name of the script that the application
monitor must run to perform notification when the application fails.
The
application monitor runs this script each time the application monitor
restarts an application if the application fails completely or if
the application falls over to the next node in the cluster. Configuring
and specifying a script to perform notification is preferred. You
cannot specify arguments for this script.
Note: Notify scripts are
run only for failure of long-running application monitors, regardless
of the Action on application failure option
that you selected.
- Cleanup script
- Specify the full path and file name of the script that the application
monitor must run to stop the application. The default cleanup script
is the stop script that is defined for the application controller
when the application controller is configured. If the application
monitor belongs to more than one application controller, specify a
cleanup script specific to this application monitor. In this case,
do not rely on the default stop script.
If the Monitor
mode field is set to Startup monitoring,
you do not have to specify a cleanup script. PowerHA SystemMirror will
not use it. You cannot specify arguments for this script.
- Restart script
- Specify the full path and file name of the script that the application
monitor must run to start the application. The default restart script
is the start script that is defined for the application controller
when the controller is configured. If the monitor belongs to more
than one application controller, there is no default stop script.
If
the Monitor mode field is set to Startup
monitoring, you need not specify a restart script because PowerHA SystemMirror will
not use it. You cannot specify arguments for this script.
- Application controllers
- Select one or more application controllers for the application
monitor to monitor.