摘要 |
Relationships between components in an application and the services they provide are identified, including redundant caller-callee sequences. Specific components of interest are instrumented to obtain data when they execute. Data structures are created which identify the components and their dependencies on one other. To avoid excessive overhead costs, redundant dependencies are identified. A dependency data structure can be provided for each unique dependency. When repeated instances of a dependency are detected, the associated dependency data structure can be augmented with correlation data of the repeated instances, such as transaction identifiers and sequence identifiers. Sequence identifiers identify an order in which a component is called. A flag can be used to force the creation of a new dependency data structure, and a calling component's name can be used instead of a sequence identifier. Agents report the dependency data structures to a manager to provide graph data in a user interface.
|