Defining Event Types

Events are the fundamental elements for information exchange in an event processing application. An event captures important information about the state (or changes to state) of some entity. Examples of events include stock market trades and quotes, ATM transactions, radio-frequency identification (RFID) signals, and telemetry readings.

The event correlator is Apama’s core event processing and correlation engine. The correlator monitors input events and produces output events.

Before you send event data to the correlator, your application needs to define the types (or structures) of events so that the correlator can interpret the events correctly when it receives them.

A typical Apama application processes streaming data coming from real-time custom or standards-based data feeds.

For the purposes of this tutorial, the sample application you build will monitor simulated temperature and pressure telemetry readings (input events) from a fictional industrial process.

In this first lesson, you will define the fields and data types for the temperature event and pressure event input from a simulated data feed.

  1. Clone the Apama Samples repository, then open the apama-samples/tutorials/EPL_Fundamentals project in VS Code.

  2. Right-click on the eventdefinitions folder and create a new file SensorMonitorEvents.mon.

Here is the syntax for an event type definition:

event <EventType> {  
  <fieldtype> <fieldName>; 
  ...  
}

The industrial process’s Temperature readings have the following event fields in this order:

  • sensorId (string)
  • temperature (float)
  1. Add the following event definition for Temperature events in the SensorMonitorEvents.mon file:
event Temperature {
    string sensorId;
    float temperature;  
}
Info
  • Event and field names are case-sensitive and fields are strongly typed. You must use the names and types shown here for use later in this tutorial.
  • Field order is also important when you define an event - the order of fields in the event type definition must match the order of fields input from the incoming events (in this case, the SensorTest.evt file).
  • The sensorId is a unique identifier for the sensors in the fictional industrial process. The field Temperature identifies the temperature in Celsius degrees. The field Pressure identifies the pressure in kilopascals.

  1. In the SensorMonitorEvents.mon file, after the definition of the Temperature event type, add a separate event definition for the Pressure event type.

  2. Save your changes. The contents of the SensorMonitorEvents.mon file should now look like this:

/**  
 * Input temperature reading.  
 */  
event Temperature {  
    string sensorId;  
    float temperature;  
}

/**  
 * Input pressure reading.  
 */  
event Pressure {  
    string sensorId;  
    float pressure;  
}
Info
The /** Comment */ pattern is the recommended way to document the purpose and usage of event definitions in EPL.
Info
The sensorId is a unique identifier for the sensors in the fictional industrial process. The targetTemperature and targetPressure fields provide the mean values for temperature and pressure.

  1. In the SensorMonitorEvents.mon file, after the definition of the Pressure event type, add a separate event definition for the AddSensor event type. The AddSensor events have the following fields in this order: sensorId (string), targetTemperature (float), and targetPressure (float).

  2. Save your changes. The contents of the SensorMonitorEvents.mon file should now look like this:

/**  
 * Input temperature reading.  
 */  
event Temperature {  
    string sensorId;  
    float temperature;  
}

/**  
 * Input pressure reading.  
 */  
event Pressure {  
    string sensorId;  
    float pressure;  
}  
  
/**  
 * Sensor configuration.
 */ 
event AddSensor {  
    string sensorId;  
    float targetTemperature;  
    float targetPressure;  
}

Now that you have added some event definitions, temperature and pressure telemetry readings (for the fictional industrial process that your sample application will monitor) are simulated via the SensorTest.evt file.

  1. Expand the events folder.

  2. Open SensorTest.evt.

  3. Look at the format of the data in this file. There are three types of events, corresponding to the events you have just defined: AddSensor, Temperature, and Pressure.

Notice how the Temperature and Pressure events are each preceded by a string (“Factory1” or “Factory2”). This optional string defines the name of the channel that the event is sent on.

Info
The BATCH lines in the .evt file send events to the correlator in groups, timed in milliseconds.
  1. Close the SensorTest.evt file.