Overview

Functional description

You can create policies to raise NFM-P alarms based on managed-object statistics. When a statistic counter of an object such as an interface reaches a specified threshold, the NFM-P alerts GUI operators using a threshold-crossing alarm, or TCA, based on policy specifications.

A TCA policy includes the following:

Note: A TCA is not self-clearing. The NFM-P clears a TCA only when the TCA policy contains a falling-threshold rule in addition to a rising-threshold rule, and the alarm severity in the falling-threshold rule is set to cleared.

TCA operation

To enable TCA on an object, you must enable performance statistics collection on the object. See the NSP NFM-P Statistics Management Guide for information about enabling performance statistics collection.

The NFM-P compares the object counter values at each statistics collection to the threshold values in the associated TCA policy. When a value initially crosses a threshold, the NFM-P raises an alarm. Rather than raise new alarms for successive threshold-crossing events associated with the policy, the NFM-P adjusts the alarm severity based on the policy rules.

Note: If required, you can use a parameter on the TCA policy properties form to disable the alarm-severity demotion defined in a TCA policy. For example, an operator may want a critical alarm associated with an object to persist and be visible to other operators until the fault condition is cleared.

Using the NFM-P client GUI, you can associate a TCA policy with a managed object from the TCA policy configuration form, or from the object properties form. You can associate one TCA policy with multiple objects.

TCA policy rules

Depending on the object type, you can use rules to monitor the following:

A rule can apply to an absolute counter value or to a delta value, which is the difference between the current value and the previous value.

In a TCA policy rule, you can specify whether the rule raises alarms. Disabling the raising of alarms according to a policy rule may be required, for example, when the rule raises excessive alarms because of a fault that causes frequent threshold crossing. You can disable the raising of alarms while the fault is being resolved, and then re-enable the raising of alarms when the affected object returns to normal operation.

Custom profile TCAs

You can optionally create custom profile TCAs that define the TCA rules using one or more of the following in a mathematical formula:

Note: For monitored object property values in a formula, the current and previous values are the same. If a monitored object property value changes between collections, the new property value is used as the current and previous values.

After you create a custom profile TCA, you associate the custom profile TCA with a TCA policy to put the TCA into effect.

TCA configuration example

The following example describes the NFM-P configuration required to raise a TCA, adjust the alarm severity, and clear the alarm, based on the following utilization specifications:

The following are the TCA rules and associated parameter values required to implement the example alarm behavior:

Rule to raise minor alarm for 60% or higher utilization:

Rule to change alarm severity to major for 70% or higher utilization:

Rule to change alarm severity to critical for 80% or higher utilization:

Rule to change alarm severity to major if utilization falls below 80%:

Rule to change alarm severity to minor if utilization falls below 70%:

Rule to clear alarm if utilization falls below 60%: