CPU Statement—Specify Conditions to Monitor CPU Use

C-LANG Statements › CPU Statement—Specify Conditions to Monitor CPU Use

CPU Statement—Specify Conditions to Monitor CPU Use

The CPU statement specifies the conditions that a CPU Monitoring job uses to monitor the CPU usage of the computer where the agent is installed.

Supported Job Type

This statement is optional for the CPU Monitoring job type.

This statement does not apply when the WAITMODE statement is set to NOW.

Syntax

This statement has the following format:

CPU [FROM(lower)] [TO(upper)] [NOCHANGE(percent)] [CONTINUOUS(alertid)] [WITHIN|OUTSIDE] [AVAILABLE|USED]

FROM(lower)

(Optional) Defines the lower boundary of CPU usage in percent.

Limits: 0-100

Note: If you specify the FROM operand without the TO operand, the range is between the lower boundary and 100 percent.

TO(upper)

(Optional) Defines the upper boundary of CPU usage in percent.

Limits: 0-100

Note: If you specify the TO operand without the FROM operand, the range is between zero percent and the upper boundary.

NOCHANGE(percent)

(Optional) Defines the percentage value that the CPU usage must change to trigger the alert. The trigger will not occur if the value change is within the specified percentage value.

Limits: 0-100

Notes:

To use the NOCHANGE operand, specify an alert in the CONTINUOUS operand.
The alert will trigger only if the following conditions are met:
- The change value is within the range specified by the FROM and TO operands.
- The change value is greater than the delta specified by the last scanned amount. That is, the first time CPU usage matches the job criteria, an alert is triggered. Subsequently, an alert is triggered only if the change in CPU usage is greater than the delta calculated using the last scanned amount that was registered in a trigger.

CONTINUOUS(alertid)

(Optional) Specifies the identifier of an alert to be triggered when the specified CPU monitoring conditions occur. Defines the job as continuous. To end continuous monitoring, you must complete the job manually or cancel it. If you do not specify this operand, the job completes when the specified CPU monitoring conditions occur.

alertid

Specifies the alert identifier to be triggered.

Limits: 1-8 alphanumeric characters; the first character must be alphabetic

Note: The alert must have been previously defined to the scheduling manager. Not all scheduling managers support the CONTINUOUS operand.

Note: When using the CONTINUOUS operand, you must also specify the FROM operand, the TO operand, or both.

WITHIN|OUTSIDE

(Optional) Specifies whether the job completes (or triggers an alert if monitoring continuously) when the value of CPU usage is within or outside the specified range.

WITHIN: Specifies that the job completes (or triggers an alert if monitoring continuously) when the value of CPU usage is within the range specified by the FROM and TO operands. This is the default.
OUTSIDE: Specifies that the job completes (or triggers an alert if monitoring continuously) when the value of CPU usage is outside the range specified by the FROM and TO operands.

AVAILABLE|USED

(Optional) Specifies whether the job monitors the available or used CPU capacity.

AVAILABLE: Specifies that the job monitors the available CPU processing capacity. This is the default.
USED: Specifies that the job monitors the used CPU processing capacity.

Note: On UNIX, the CPU calculations are based on load averaging as displayed by the command top. You can scale the results the agent returns with the objmon.cpu.scalefactor parameter. For more information about the objmon.cpu.scalefactor parameter, see the CA Workload Automation Agent for UNIX, Linux, or Windows Implementation Guide.
You must specify at least one of the operands FROM or TO in the CPU statement when using the WAITMODE statement set to the WAIT value.

The CPU statement does not apply when using the WAITMODE statement set to the NOW value.

Example: Continuously Monitor and Report CPU Usage Outside a Range

This example continuously monitors the CPU available on the SYSAGENT computer. An alert is triggered each time the available CPU is less than 20 percent or greater than 50 percent, and the CPU usage changes by more than 10 percent. If the percent change is less than or equal to 10 percent, the job does not register a change.

AGENT SYSAGENT
CPU FROM(20) TO(50) NOCHANGE(10) CONTINUOUS(al3) +
  OUTSIDE AVAILABLE