A health rule violation occurs when the critical or warning condition set for that health rule evaluates as true. In this case, a Violation Started - Critical event or a Violation Started - Warning event generates a health rule violation with an Open status.

A health rule violation ends when it is either:

  • Resolved: The reported metrics indicate that the violated condition is no longer true.
    A Violation Created - Critical event or a Violation Ended - Warning to Normal or Violation Ended - Critical to Normal event generates when the violation status resolves.

or

  • Canceled: The health violation status of the health rule cannot be accurately determined if it continues to violate or has ended. In this case, the health rule violation status is Canceled. This can occur when:
    • The health rule is edited.
    • The health rule is disabled.e
    • Health rule evaluation entities are added or removed.
    • Health rule violation metric values have become UNKNOWN.

A Violation Ended or Violation Canceled event generates when the violation status of a health rule changes to Canceled.

A new health rule violation starts if the same health rule violates after a resolution or cancellation of a previous violation. 

During the life of one health rule violation, there may be other types of health rule violation events, such as Health Rule Violation Ungraded/Downgraded/Continues events.

The following figure illustrates the health rule violation life cycle:

Health Rule Violation Lifecycle

View Health Violation Timeline

You can view the health violation details for a selected entity or any health rule in the health violation timeline. To access the health violation timeline:

  1. In the Relationships panel, select any entity. A list view of all the instances and the associated details of that entity appears.
  2. Click on the name of any entity instance. The entity performance details view replaces the list view displaying the health violation details and other performance details over a time period.
    Health Rules List for an Entity Type

    Health Violation Details for a Selected Entity
  3. Expand the Health Violations (n) section to view a list of all violating health rules associated with the selected entity. 

View Violating Metrics

On the Health Violations (n) section, click a health rule to view the associated violating metrics. The Violating Metrics chart provides a graphical representation of the baselines and the threshold values that help you to quickly examine the health violation details. For health rules where single metric thresholds are defined, you can view the threshold value on the chart. For health rules where baselines are used, you can view the baseline, the mean value, the upper band, and the lower band.

If standard deviation is selected, the upper band and lower band of baseline are calculated as follows:

  • Upper Band = Mean Value + (Factor * Standard Deviation)
  • Lower Band = Mean Value - (Factor * Standard Deviation)

If the percentage deviation is selected, the upper band and lower band of baseline are calculated as follows:

  • Upper Band = Mean Value + (Factor * Mean Value/ 100)
  • Lower Band = Mean Value - (Factor * Mean Value/ 100)

Anomaly Detection

AppDynamics Cloud Anomaly Detection automatically determines whether every service in your application performs within the acceptable performance limits. It uses machine learning capabilities to continuously monitor latency, errors, and throughput of services to identify abnormal behavior. This helps reduce the Mean Time To Detection (MTTD) for application performance problems.

From Observe > Application Performance Monitoring > Services, select a service to view the detected anomaly details for that service. The color-coded status symbol next to Anomaly Detection in the Health and Alerting section depicts the overall status of the service:

  • Critical Health Status—Indicates the detected anomalies are in critical state. You may want to take corrective actions. 
  • Unknown Health Status—Indicates the detected anomalies are in unknown state. An anomaly is rendered to unknown state when there is no enough data to deterministically evaluate it.
  • Warning Health Status—Indicates the detected anomalies are in warning state.
  • No indicator implies that the service is performing within the expected range.

You can view the relationship between the selected service and other services as a flow map or a list. 

Anomaly Detection is enabled by default for all services in the application. It takes 48 hours for machine learning to model train on the services in your application. Click Configure > Anomaly Detection to view the model training status and other details of all the services. See Model Training.

Once the model training is complete, you can view the detected anomalies, monitor them, and take corrective actions. See Monitor Anomalies.

Examine the Health Violation

You can view data related to the violation in the Health Violation Timeline and the Health and Alerting sections in the right panel. This data helps pinpoint the exact condition or metric that is violating and take corrective actions. 

The Health Violation Timeline displays a list of violating health rules associated with the selected entity along with the Critical or Warning status. Red indicates Critical status while Yellow indicates Warning status. The health rule name displays above the timeline. The violation timeline and the start time of the health violation appear at the bottom. The end time of the health violation displays depending on the data collection duration you choose. 

To view more details, select the health rule. The following Violation Details display in the right panel:

  • Health rule name
  • Violating metric
  • Entity name
  • Average value of the violating metric
  • Start date and time
  • On going violation duration
  • Status of the violation

The Health Violation Timeline displays the anomaly detection status details if you select a service.

Refer to the following image to view the numbered details of violation data. 
Examine Health Violation

How is the Health of an Entity with Multiple Health Rules Determined?

You can configure multiple health rules to monitor an entity. If you configure multiple health rules for an entity, the worst performing health rule determines the health of the entity

View Details of Violating Health Rule

The Health Rules button on the Health and Alerting section in the right panel displays the overall status of the selected entity. This Critical Health Status indicates that the entity is not healthy and one or more health rules associated with the entity are violating. To view the health status of all health rules, click Health Rules. A list of all health rules associated with the entity appears along with the health status. The worst-performing health rule determines the health status of an entity with multiple health rules. The following icons depict the health status:

  • Critical Health Status—Indicates the health rule is violating and is in a critical state. You may want to take corrective actions. 
  • Unknown Health Status—Indicates the health rule is in unknown state. A health rule is rendered to unknown state when there is no enough data to deterministically evaluate the conditions.
  • Warning Health Status—Indicates the health rule is violating and is in a warning state.
  • Normal Health Status—Indicates that the monitored metric is performing within the expected range.

Violating Health Rule Details

This list also presents other health rule details such as:

  • The number of monitored objects.
  • The health rule evaluation is enabled.
  • Actions linked to the health rule.

Enable/Disable Health Rules

You can enable or disable the evaluation of the health rules in the health rules list. You can enable or disable the evaluation of a single health rule or all the health rules associated with the entity. Once you disable a health rule, the evaluation of that health rule is suspended until you re-enable it again.

A health rule does not appear on the list of health rules on the <entity> details page if it is disabled. View the disabled health rules on the Health Rules list.

View Status of Monitored Objects

Click the Applied on 'n' objects associated with the health rule to view a list of all monitored objects. The health status of each monitored object displays next to the object.
Monitored Entities

Add/Update Actions Triggered by a Health Violation

Click the Link action associated with the health rule. The Edit Health Rule wizard appears. Add or update actions as required. These actions trigger when the health rule violates. See Edit a Health Rule.

You must create an action before you link it to the health rule. See Actions.