Build and Deploy Streaming Lite Project

1. Create project "freezer_monitoring_lite"

   In HANA Studio, create a new HANA Streaming Analytics project called "freezer_monitoring_lite", in the “default” workspace. For detailed instructions on creating a project, you can refer back to the Streaming Analytics: Run and Test a Streaming Project tutorial.

2. Copy CCL code

   Copy the following CCL code into our "freezer_monitoring_lite" project. Remember to delete the NEWSTREAM element that is created by default in each new project.

   
   CREATE SCHEMA FreezerTemperatureReading (
   	SensorId string ,
   	Temperature float,
   	ReadingDate msdate,
   	Id long) ;
   
   CREATE INPUT STREAM isFreezerTemperatureReading
   SCHEMA FreezerTemperatureReading AUTOGENERATE ( Id ) ;
   
   /**@SIMPLEQUERY=AGGREGATE*/
   CREATE OUTPUT WINDOW aAverageTemperatureBySensor
   PRIMARY KEY DEDUCED
   AS
   	SELECT
   		isFreezerTemperatureReading.SensorId MACHINEID ,
   		last (isFreezerTemperatureReading.ReadingDate) EVENT_TIME ,
   		'TEMP'EVENT_NAME ,
   		'' EVENT_DESCRIPTION ,
   		string ( avg ( isFreezerTemperatureReading.Temperature ) )
   EVENT_VALUE ,
   		count (isFreezerTemperatureReading.Temperature) rowct  
   FROM isFreezerTemperatureReading KEEP EVERY 15 ROWS PER (SensorId)
   GROUP BY isFreezerTemperatureReading.SensorId ;
   
   /**@SIMPLEQUERY=FILTER*/
   CREATE OUTPUT STREAM FilterEveryFifteenthRow
   AS SELECT
   	aAverageTemperatureBySensor.MACHINEID,
   	aAverageTemperatureBySensor.EVENT_TIME,
   	aAverageTemperatureBySensor.EVENT_NAME,
   	aAverageTemperatureBySensor.EVENT_DESCRIPTION,
   	aAverageTemperatureBySensor.EVENT_VALUE
   FROM aAverageTemperatureBySensor
   WHERE	aAverageTemperatureBySensor.rowct = 15 ;
   

   

   The elements used in this project are:

   • isFreezerTemperatureReading

     Input stream for temperature sensor readings. In this tutorial, we will be using the streamingupload command line utility to write rows of data to this stream.

   • aAverageTemperatureBySensor

     • Incoming temperature values are grouped by SensorId

     • The average temperature is calculated for each SensorId based on the temperature values currently in the window for the given SensorId.

     • KEEP EVERY 15 ROWS PER (SensorId) policy specifies after receiving and processing the 15th value for a given SensorId, all values for that SensorId are cleared from the window and the count is reset to 0. This behavior is referred to as a “jumping” window. If the KEEP clause did not use the EVERY keyword, then it would be a sliding window in which the arrival of the nth value would push out the n-15th value. (The arrival of the 16th value would push out the 1st value, the 17th value would push out the 2nd value)

     • There is a counter to count the number of rows currently in the window

     • Maps the input schema into the schema that our SDS project freezer_monitoring uses

   • FilterEveryFifteenthRow

     • This filter allows a row through when the aAverageTemperatureBySensor counter equals 15 for a given SensorId

     • Result is that for every 15 rows that enter the project for a given SensorId, one row is output from the filter for that SensorId

3. Compile

   After creating our streaming project, compile it to generate the .ccx file. It will be located in the bin/ directory of our project folder.

   

For the question below, select the correct answer, and click Validate.

Connect to your Raspberry Pi with WinSCP, and move your .ccx file over onto the Pi. We are moving it into the /home/pi directory.

Run the Streaming Lite project using the command line. You can do this directly from a shell session on the Pi or remotely using Putty to open an SSH session. In either case you will run the following command line:

$HOME/install/bin/streamingproject --ccx $HOME/freezer_monitoring_lite.ccx --command-port 9230

The first path points to streamingproject, a binary executable from the Streaming Lite “install” folder. The second path points to the freezer_monitoring_lite.ccx project file. We then specify the port which our project will run on.

You should now see our Streaming Lite project running: