Runtime metrics collection processing is an offline utility that processes resource usage information that is collected by using name-value pair collection and writes the processed results to a relational database. Analysis packages, for example, an SQL dashboard, can be used to process and visualize the results. Sample SQL procedures are included in this APAR.
With name-value pair collection and runtime metrics collection, you can understand and analyze system resource usage by horizontal or vertical annotations, or both, in your application. Horizontal annotations are made in your application by using name-value pairs and can represent various aspects, such as message type, message origin, customer, input values, and so on. Vertical annotations are made in your application by using ECB owner names, and can represent code packages and phases of processing. The data is collected on the z/TPF system and runtime metric collection processing writes the processed data to a relational database.
With name-value pair collection and runtime metrics collection:
• A coverage programmer can gain new insights into system resource usage and identify the source of problems considerably faster than previously required.
• A capacity planner can determine additional physical assets that are required based on new resource metrics and the message growth rate of a specific service.
• An application architect can identify inefficient code packages that can be refactored to improve system performance.
With a minimal investment, you can realize a tremendous amount of value from name-value pairs by using runtime metrics collection.