Model Types

There are three types of models:

  • Absolute Models

  • Relative Models

  • Instantiable Models

Absolute Models

An absolute model has a constant data source. It uses absolute references to pull data from different parts of the system. All variables, functions, and events declared by an absolute model are added to the model's own context.

For example, a model simulating a technological process could take data from multiple specific controllers and perform complex mathematical processing over them. Results of calculations would be visible as model properties. Such a model would be configured as an absolute model.

An absolute model's bindings and rule expressions cannot include relative references.

Relative Models

Relative models are designed to process data from a certain source context and its children. A separate instance of a relative model is created for each context it is attached to.

For example, a model calculating the CPU load of a network device with respect to its type and communication should attach to every network device, apply a rule set to its properties, and add a "generic" CPU Load property to each device. Such model should be configured as a relative model.

A relative model's bindings and rule expressions can include relative references.

Instantiable Models

Instantiable models have instances that are created on demand. Once you define an instantiable model, Iotellect Server creates several containers for storing model instances and attaches those containers to different places of Iotellect Server context tree.

Thus, instances of instantiable models are very similar to other Iotellect Server objects, such as Alerts or Reports. However, those instances are fully customized. Their properties and behaviour is described by the model.

For example, you may be managing oil derricks. Each derrick is a complex object that involves a number of devices providing diverse telemetry and a number of server-side rules for storing and processing statistics, preparing printable reports, etc.

In this case, you'll need to create an Oil Derrick instantiable model and attach containers of this model's instances to Iotellect Server user accounts (in exactly the same way as other objects are owned by users).

Each instance of the model will refer to specific devices providing the necessary telemetry data, as well as all necessary variables/functions/events, bindings, and business rules required for processing and visualizing collected data.

See Using Instantiable Models for more information.

Was this page helpful?