Root Cause Analysis (RCA) processing is triggered by the occurrence of a service failure event. Once an event occurs RCA evaluates the state of the service and system components on which the service depends. RCA considers the presence of other performance and availability events for these dependent components as well as collecting the necessary data from the Enterprise Manager repository and Agents.
RCA considers the status (availability) of the system components and any component tests you have defined. Component tests are used to evaluate key aspects of the components on which the service depends but which Enterprise Manager does not recognize as managed targets.
For example if your service depends on a database, the service may fail if a tablespace used by your service is offline. In this situation the service could fail even though the database itself is still available. You could add a component test to the database that tests for the tablespace being online. When RCA runs, it evaluates the status of key component and component tests and reports violations as possible root causes. A component may have different component tests for each service that depends on it; the tests are specific to a service.
RCA also includes the analysis results with notifications sent for service failure alerts. When you register for notification of service failures, you may choose to wait for RCA to complete and have the results of the analysis included in the notification sent to you when the service fails.
RCA continues to analyze the service failure as long as it persists, periodically checking for changes to the status of the system components and any component tests that you have defined. If the nature of the service failure changes over time, RCA updates its analysis to reflect any different causes of the failure.
Related Topics
Getting the Most From Root Cause Analysis
Enterprise Manager Information Roadmap
Copyright © 1996, 2005, Oracle. All rights reserved.