Analyzing CDAP Data from BI Tools

Source Code Repository: Source code (and other resources) for this guide are available at the CDAP Guides GitHub repository.

Cask Data Application Platform (CDAP) provides an abstraction— Datasets —to store data. In this guide, you will learn how to integrate and analyze Datasets with BI (Business Intelligence) Tools.

What You Will Build

This guide will take you through building a CDAP application that processes purchase events from a Stream, persists the results in a Dataset, and then analyzes them using a BI tool. You will:

  • build a CDAP Application that consumes purchase events from a Stream and stores them in a Dataset;
  • build a Flowlet that processes purchase events in realtime, writing the events into a Dataset; and
  • finally, access this Dataset from a BI tool to run queries by joining purchase events in the Dataset with a product catalog—a local data source in the BI tool.

Let’s Build It!

The following sections will guide you through building an application from scratch. If you are interested in deploying and running the application right away, you can clone its source code and binaries from this GitHub repository. In that case, feel free to skip the next two sections and jump right to the Build and Run Application section.

Application Design

In this example, we will learn how to explore purchase events using the Pentaho BI Tool. We can ask questions such as "What is the total spend of a customer for a given day?"


A purchase event consists of:

  • Customer
  • Quantity purchased
  • Product

Purchase events are injected into the purchases Stream. The sink Flowlet reads events from the Stream and writes them into the PurchasesDataset. The PurchasesDataset has Hive integration enabled and can be queried, like any regular Database table, from a BI tool by using the CDAP JDBC Driver.


The first step is to get our application structure set up. We will use a standard Maven project structure for all of the source code files:


The application is identified by the PurchaseApp class. This class extends AbstractApplication, and overrides the configure() method to define all of the application components:

public class PurchaseApp extends AbstractApplication {

  public static final String APP_NAME = "PurchaseApp";

  public void configure() {
    addStream(new Stream("purchases"));
    addFlow(new PurchaseFlow());
    createDataset("PurchasesDataset", PurchaseStore.class,;

When it comes to handling time-based events, we need a place to receive and process the events themselves. CDAP provides a real-time stream processing system that is a great match for handling event streams. After first setting the application name, our PurchaseApp adds a new Stream, called purchases.

We also need a place to store the purchase event records that we receive; PurchaseApp next creates a Dataset to store the processed data. PurchaseApp uses an ObjectStore Dataset to store the purchase events. The purchase events are represented as a Java class, Purchase:

public class Purchase {
  private final String customer;
  private final String product;
  private final int quantity;
  private final long purchaseTime;

  public Purchase(String customer, String product, int quantity, long purchaseTime) {
    this.customer = customer;
    this.product = product;
    this.quantity = quantity;
    this.purchaseTime = purchaseTime;

  public String getCustomer() {
    return customer;

  public long getPurchaseTime() {
    return purchaseTime;

  public int getQuantity() {
    return quantity;

  public String getProduct() {
    return product;

  public byte[] getKey() {
    String hashedKey = purchaseTime + customer + product;
    return hashedKey.getBytes();

PurchaseApp adds a PurchaseFlow to process data from the Stream and store it into the Dataset:

public class PurchaseFlow extends AbstractFlow {

  public void configure() {
    setDescription("Reads purchase events from a stream and stores the purchases in a Dataset");
    addFlowlet("sink", new PurchaseSinkFlowlet());
    connectStream("purchases", "sink");

The PurchaseFlow contains a PurchaseSinkFlowlet that writes to the Dataset:

public class PurchaseSinkFlowlet extends AbstractFlowlet {

  private static final Logger LOG = LoggerFactory.getLogger(PurchaseSinkFlowlet.class);

  private PurchaseStore store;

  public void process(StreamEvent event) {
    String body = Charsets.UTF_8.decode(event.getBody()).toString();
    // <customer>,<quantity>,<productId>
    String[] tokens =  body.split(",");
    for (int i = 0; i < tokens.length; i++) {
      tokens[i] = tokens[i].trim();
    if (tokens.length != 3) {
      LOG.error("Invalid stream event:{}", body);
    String customer = tokens[0];
    int quantity = Integer.parseInt(tokens[1]);
    String item = tokens[2];

    Purchase purchase = new Purchase(customer, item, quantity, System.currentTimeMillis());

PurchaseStore is a custom dataset that implements the RecordScannable interface for integration with Hive:

public class PurchaseStore extends AbstractDataset implements RecordScannable<Purchase> {

  private final ObjectStore<Purchase> store;

  public static DatasetProperties properties() {
    try {
      return ObjectStores.objectStoreProperties(Purchase.class, DatasetProperties.EMPTY);
    } catch (UnsupportedTypeException e) {
      throw new RuntimeException("This should never be thrown - Purchase is a supported type", e);

  public PurchaseStore(DatasetSpecification spec,
                       @EmbeddedDataset("store") ObjectStore<Purchase> objStore) {
    super(spec.getName(), objStore); = objStore;

  public Type getRecordType() {
    return Purchase.class;

  public List<Split> getSplits() {
    return store.getSplits();

  public RecordScanner<Purchase> createSplitRecordScanner(Split split) {
    return Scannables.valueRecordScanner(store.createSplitReader(split));

  public void write(Purchase purchase) {
    store.write(purchase.getKey(), purchase);

  public Purchase read(byte[] key) {

Build and Run Application

The PurchaseApp application can be built and packaged using the Apache Maven command from the project directory:

$ mvn clean package

Note that the remaining commands assume that the cdap script is available on your PATH. If this is not the case, please add it:

$ export PATH=$PATH:<CDAP home>/bin

If you haven't already started a standalone CDAP installation, start it with the command:

$ cdap sdk start

We can then deploy the application to a running standalone CDAP installation:

$ cdap cli load artifact target/cdap-bi-guide-<version>.jar
$ cdap cli create app PurchaseApp cdap-bi-guide <version> user
$ cdap cli start flow PurchaseApp.PurchaseFlow

Next, we will send some sample purchase events into the stream for processing. The purchase event consists of a customer name, a quantity purchased and a product purchased:

$ cdap cli send stream purchases \"Tom,    5,       pear\"
$ cdap cli send stream purchases \"Alice, 12,      apple\"
$ cdap cli send stream purchases \"Alice,  6,     banana\"
$ cdap cli send stream purchases \"Bob,    2,     orange\"
$ cdap cli send stream purchases \"Bob,    1, watermelon\"
$ cdap cli send stream purchases \"Bob,   10,      apple\"

Now that purchase events have been ingested by CDAP, they can be explored with a BI tool such as Pentaho Data Integration.

  1. Download Pentaho Data Integration and unzip it.

  2. Before opening the Pentaho Data Integration application, copy the <CDAP home>/lib/co.cask.cdap.cdap-explore-jdbc-<version>.jar file from the CDAP SDK to the <data-integration-dir>/lib directory.

  3. Run Pentaho Data Integration by invoking <data-integration-dir>/ from a terminal.

  4. Open <cdap-bi-guide-dir>/resources/total_spend_per_user.ktr using File -> Open URL

    This is a Kettle Transformation file exported from Pentaho Data Integration. This file contains a transformation that calculates the total spend of a customer based on the previous purchase events. The transformation has several components or steps:

    • CDAP Purchases Dataset is a step which uses the PurchasesDataset as an input source. It pulls all of the stored purchase events from CDAP.
    • The Product Catalog CSV step is another source of data, which pulls in a table from a locally defined csv file. This table contains a mapping of product name to product price, so that we can put a pricing on the purchase events.
    • The Join Rows step joins the two data sources on the product column, hence adding price information to the purchase event.
    • We use the Product Cost Calculator step to multiply purchase.quantity by price to get the total cost for the purchase.
    • The Sort on Customer sorts all of the rows by customer so that the next step can aggregate on price.
    • The Aggregate by Customer groups the rows by customer and aggregates on the total cost per purchase. This results in a table that is a mapping from customer name to a total amount spent by that customer.
  5. Double click on the CSV file input step, and change the filename to point to <cdap-bi-guide-dir>/resources/prices.csv:

  6. To run this transformation, click Action -> Run -> Launch.

  7. Once the transformation has completed execution, click on the Aggregate by Customer step, and then click on the Preview Data tab at the bottom to view the total amount spent by each customer.


Congratulations! You have now learned how to analyze CDAP Datasets from a BI tool. Please continue to experiment and extend this sample application.

Extend This Example

Now that you know how to integrate CDAP Datasets with BI Tools for data analysis, you can ask questions such as:

  • How much revenue does a particular product generate in a day?
  • What are the three most popular products?

If you were to add a ZIP code to the initial purchase events, you can then ask location-based questions such as:

  • What are the popular products in any location?
  • Which locations have the highest revenue?

Share and Discuss!

Have a question? Discuss at the CDAP User Mailing List.


Copyright © 2014-2017 Cask Data, Inc.

Licensed under the Apache License, Version 2.0 (the "License"); you may not use this file except in compliance with the License. You may obtain a copy of the License at

Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.