Introduction to Custom C/C++ Adapters for Freezer Monitoring Lite

This tutorial will go over how to create a Custom C/C++ Adapter that will work with either HANA Streaming Analytics or Streaming Lite. In the testing section, the Custom C/C++ Adapter will be executed on a remote device, and will send data into a Streaming Lite project running on that same device. It is up to the Custom C/C++ Adapter to poll any available hardware sensors for values.

The Custom C/C++ Adapter will be created using the C/C++ SDK, and built using a makefile which we will provide. In this document, we will be building an adapter to feed data into the Freezer Monitoring Lite project. The purpose of this tutorial is to show how to create your own Custom C/C++ Adapter given a specific schema.

The C/C++ Adapter fits into a larger architecture. After sending data into Streaming Lite, a Streaming Web Service output adapter carries the manipulated data into a Streaming Analytics project via the Streaming Web Service. For this tutorial, we will only be focusing on the Custom C/C++ Adapter and nothing else.

The remote device we will be using is a Raspberry Pi, running on Raspbian 7.8. However, any device which runs Streaming Lite will work for this walk-through. On the Raspberry Pi, the home directory is /home/pi, and the $STREAMING_HOME variable is our Streaming Lite installation folder, /home/pi/install.

Note: If you would like to add the Streaming Web Service Output Adapter to your Streaming Lite project first, complete these two tutorials: [Streaming Web Service Configuration and Testing (for SAP HANA Streaming Analytics)], [Streaming Web Service Output Adapter Tutorial (for Streaming Lite)]. Completion of these tutorials is not necessary for the purpose of this tutorial.

We will be developing a C/C++ project consisting of one .cpp file. This will ultimately be built on the remote device into an executable. This setup makes debugging tricky, since the C and C++ SDK used for the C/C++ Adapter will be located on the remote device. Although there are a variety of ways to set up a more suitable environment, for simplicity we will be coding directly on the Raspberry Pi, using Notepad++. Follow these instructions, and any errors can be fixed when the code is built using our Makefile.

The project we are creating this Java Adapter for is called freezer_monitoring_lite, and has been created in previous tutorial [Streaming Lite: Freezer Monitoring Lite]. The specific stream we are connecting to is called isFreezerTemperatureReading. Here is the CCL for the stream:

CREATE INPUT STREAM isFreezerTemperatureReading SCHEMA (
  SensorId string ,
  Temperature float ,
  ReadingDate msdate ,
  Id long ) AUTOGENERATE ( Id ) ;

You do not require the "SWS_Output1" Streaming Web Service Output Adapter to go through this tutorial. All you need is the isFreezerTemperatureReading input stream..

CC=g++
CPPFLAGS=-I$(STREAMING_HOME)/include
LDFLAGS=-Wl,-R,$(STREAMING_HOME)/lib –L$(STREAMING_HOME)/lib –lstreamingclient_lib –lstreamingproject_lib

custom_c_adapter: custom_c_adapter.o

clean:
  rm –f *.o custom_c_adapter