Sending Notifications Through Published Events Using Spark

Introduction

This sample demonstrates how you can send notifications through events published from WSO2 DAS using Apache Spark. The notifications are sent to alert about records of an existing table in the Spark environment satisfying a defined condition(s). This sample includes creating a table with a few product names and quantities in the DAL, and sending notifications when the quantity of a product falls below a defined value.  

Prerequisites

Set up the general prerequisites required for WSO2 DAS.

Building the sample

Follow the steps below to build the sample.

Creating the receiving event stream and the table in DAL

Follow the steps below to create a table named PRODUCTS together with the receiving event stream in the Data Access Layer (DAL).

1. Log in to the DAS management console using the following URL: https://<DAS_HOST>:<DAS_PORT>/carbon/
2. Click Main, and then click Streams.
3. Click Add Event Stream.
4. Enter the values as shown below to create an event stream named PRODUCTS_STREAM with two attributes as product name and quantity. For more information on creating event streams, see Event Streams.
   
5. Click Next (Persist Event).
6. Enter the values as shown below in the next screen to persist the created event stream. For more information on creating event streams, see  Persisting  Event Streams .
   
7. Click Save Event Stream.

Sending events to the receiving event stream

Follow the steps below to simulate the sending of events to the created receiving event stream.

1. Log in to the DAS management console using the following URL, if you are not already logged in: https://<DAS_HOST>:<DAS_PORT>/carbon/
2. Click Tools, and then click Event Simulator.
3. Upload the events to be sent to it in a CSV file (e.g. footwear.csv), and click Configure as shown below.
   
4. Enter the details as shown below, and click Configure. 
   
5. Click Play in the next screen as shown below.
   
6. Click Main, and then click Message Console and select the name of the table to view it as shown below.
   

Creating the corresponding table in Apache Spark

Follow the steps below to create a virtual table in the Apache Spark environment within WSO2 DAS to map the PRODUCTS_STREAM table in the DAL.

1. Log in to the DAS management console using the following URL, if you are not already logged in: https://<DAS_HOST>:<DAS_PORT>/carbon/
2. Click Main, and then click Console.

3. Enter the following Spark SQL query in the Spark console, and press Enter key to execute it.

   CREATE TEMPORARY TABLE PRODUCTS_MASTER 
   USING CarbonAnalytics 
   OPTIONS (tableName "PRODUCTS_STREAM", 
            schema "product-name STRING,quantity INT");

Configuring a publisher

You can attach an event publisher such as email or JMS to the PRODUCT_ALERTS_STREAM, and get the events delivered to a preferred location. Follow the steps below to configure a publisher to publish output events from WSO2 DAS, and to send notifications as logs of the terminal using a logger publisher. For more information on configuring event publishers, see Creating Alerts.

Executing the sample

Follow the steps below to execute the sample.

Creating the event stream to be published and the corresponding table in Spark

Follow the steps below to create the publishing event stream named PRODUCT_ALERTS_STREAM, and the corresponding virtual table named PRODUCT_ALERTS in the Apache Spark environment within WSO2 DAS.

1.   Log in to the DAS management console using the following URL, if you are not already logged in: https://<DAS_HOST>:<DAS_PORT>/carbon/
2. Click Main, and then click Console.

3. Enter the following Spark SQL query in the Spark console, and press Enter key to execute it.

   CREATE TEMPORARY TABLE PRODUCT_ALERTS 
   USING org.wso2.carbon.analytics.spark.event.EventStreamProvider 
   OPTIONS (streamName "PRODUCT_ALERTS_STREAM", 
             version "1.0.0",
             description "Events are published  when product quantity goes beyond a certain level",
             nickName "product alerts",
             payload "product-name STRING,quantity INT"
   );

Follow the steps below to publish output events to the created event stream.

1.  Log in to the DAS management console using the following URL, if you are not already logged in: https://<DAS_HOST>:<DAS_PORT>/carbon/
2. Click Main, and then click Console.

3. Enter the following Spark SQL query in the Spark console, and press Enter key to execute it.

   INSERT OVERWRITE TABLE PRODUCT_ALERTS  
   select * from PRODUCTS_MASTER
   where quantity<5;