What are monitor models?

A monitor model describes metrics and key performance indicators (KPIs) (collectively called business measures), their dependencies on incoming events, conditions warranting business actions (business situations), and outbound events that report these conditions and might trigger business actions.

Specifically, the monitor model describes how to perform these steps:
  • Gather information from events that will be stored in a data warehouse for reporting.
  • Group events about the same monitored entity.
  • Structure this information (for example, to allow dimensional analysis).
  • Combine this information (for example, to identify trends).
  • Visualize this information using Scalable Vector Graphics (SVG) diagrams.
  • Identify business situations in near real time, and trigger resulting actions by sending out events.

The monitor model defines the steps in enough detail that they can be executed in a runtime environment for business monitoring, such as Business Monitor.

The key to having a successful set of business measures is deciding on the few vital measures that are linked to your success. First, determine your business goals and the business measures with which you can compare your current performance with these goals. These business measures might include financial data, time-related data, throughput measures, or other performance indicators. Next, determine the business events that contain the information required to calculate the business measures. They can come from business applications, process engines, or other systems. These events often include order entry notifications, delivery notifications, and events reporting warehouse inventory transactions. Then create a monitor model that includes subscriptions to the events. Add instructions to the monitor model to describe how to calculate the business measures from event data. Based on this model, Business Monitor then provides the required information both in real time and historically.

In Business Monitor, monitor models are represented as Extensible Markup Language (XML) documents that specify how values should be extracted from events at run time and collected, combined, and stored for representation on a dashboard. The monitor model is a top-level container that contains five parts: the monitor details model (monitoring contexts, metrics, triggers, and so on), the KPI model (KPI contexts and KPIs with their events and triggers), the dimensional model (cubes, dimensions, and measures), the visual model (diagrams associated with monitoring contexts and KPI contexts for display in the dashboard), and the event model (references to event definitions for all inbound and outbound events).

This section briefly explains the following parts of the monitor model container: