Runtime Architecture

TIBCO BusinessEvents Views adds the dashboard agent to the cluster and adds additional functionality to the inference agents.

Without TIBCO BusinessEvents Views, a typical TIBCO BusinessEvents cluster consists of these types of agents:

  • Inference agents listen for events and perform inferencing on them using forward-chaining rules.
  • Cache agents store the in-memory cache data.

TIBCO BusinessEvents Views adds Dashboard agents to the cluster:

  • Dashboard Agent serves as a presentation server. It runs queries against the cache or JDBC backing store (depending on configuration). It listens for metric value updates and presents them to concurrent subscribed users in the form of dashboard components (charts and state model components).

TIBCO BusinessEvents Views also adds functionality to inference agents. Inference agents measure and compute incoming data using Metric resources. The computed data (and related information if any) are written to the cache and JDBC backing store (depending on configuration). Dashboard agents query this data and render it visually in charts on TIBCO BusinessEvents Views Dashboard.

Behavior

Behavior summary of what happens at runtime is explained in the following table.

Behavior summary
Component Summary
Inference agents At runtime, events receive data and trigger rules to create metrics with that data (combining it with other data as desired).
Cache and backing store Metric instances are written to the cache or JDBC backing store (depending on configuration) and are updated each time new data is received and processed by the rules. Historical (and other) data can be saved to a tracking table (if that feature is enabled).
Dashboard agents Data sources (queries) in dashboard agents query the metric data in the cache or JDBC backing store (depending on configuration) and provide filtered data to Dashboard clients.
Dashboard Dashboard represents the data graphically in dashboard components. User roles determine access to the dashboard components.