About Java Server Health
By default, AppDynamics provides predefined rules for CPU utilization, physical memory utilization, JVM heap utilization, and CLR heap utilization. For example the default health rule for CPU utilization triggers a warning when a node exceeds 75% CPU utilization and triggers a critical event when CPU utilization is 90% or above.
Node health is driven by node health rules. The following example shows that Node_8003 is experiencing a JVM heap health rule violation; all of the heap is consumed.
You can view the health rule violation details and the status of the violation:
There are many types of health rules and each defines a condition or set of conditions in terms of a certain set of metrics that serve as health indicators of your application component. For example, the Business Transaction Performance health rule defines a set of conditions in terms of business transaction metrics, while the Node Health-Hardware,JVM,CLR health rule defines a set of conditions in terms of hardware metrics.
To change or add a new health rule, see Configure Health Rules.
You can control the scope of a health rule to a specific application component. For example, for Node Health, you can choose between scoping to tiers or nodes. For tiers, you can apply the health rul to all tiers or to specific tiers only. For nodes, you can apply the health rule to all nodes. If you have a large cluster, you can choose specific nodes as well.
Once you scope the health rule, you can define the triggering conditions of the health rule. There are two status condition sets, warning and critical, which can be defined independently of each other. Each status condition set consists of atomic conditions that must either be ALL met or have ANY that are met in order to trigger the status condition. Atomic conditions are based on predefined metrics that serve as health indicators of your application component. If you cannot find a Health Rule type that has a predefined metric that meets your needs, you can add a metric using custom monitors or create a JMX metric from MBeans and use a Custom Health Rule.
After defining Health Rules, you can use the Health Rule Violation Events in policies to trigger actions that can notify Administrators via email or SMS systems or perform some other action.