Spark Fundamentals

Expand all

Course Description

Success of many organizations depends on their ability to derive business insights from massive amount of raw data coming from various sources.

This Spark Fundamentals training course offers many engineering improvements over the traditional MapReduce programming model as implemented in Hadoop by providing multi-pass in-memory processing of data which boosts the overall performance of your ETL and machine-learning algorithms.

Objectives

This high-octane Spark training course provides theoretical and technical aspects of Spark programming.  The course teaches developers Spark fundamentals, APIs, common programming idioms and more.

This Spark Fundamentals training course is supplemented by hands-on labs that help attendees reinforce their theoretical knowledge of the learned material and quickly get them up to speed on using Spark for data exploration.

Topics

•     Elements of functional programming
•     Spark Shell
•     RDDs
•     Parallel processing in Spark
•     Spark SQL
•     ETL with Spark
•     MLib Machine Learning Library
•     Graph Processing with GraphX
•     Spark Streaming

Course Outline

Chapter 1. Introduction to Functional Programming

•     What is Functional Programming (FP)?
•     Terminology: First-Class and Higher-Order Functions
•     Terminology: Lambda vs Closure
•     A Short List of Languages that Support FP
•     FP with Java
•     FP With JavaScript
•     Imperative Programming in JavaScript
•     The JavaScript map (FP) Example
•     The JavaScript reduce (FP) Example
•     Using reduce to Flatten an Array of Arrays (FP) Example
•     The JavaScript filter (FP) Example
•     Common High-Order Functions in Python
•     Common High-Order Functions in Scala
•     Elements of FP in R

Chapter 2. Introduction to Apache Spark

•     What is Spark
•     A Short History of Spark
•     Where to Get Spark?
•     The Spark Platform
•     Spark Logo
•     Common Spark Use Cases
•     Languages Supported by Spark
•     Running Spark on a Cluster
•     The Driver Process
•     Spark Applications
•     Spark Shell
•     The spark-submit Tool
•     The spark-submit Tool Configuration
•     The Executor and Worker Processes
•     The Spark Application Architecture
•     Interfaces with Data Storage Systems
•     Limitations of Hadoop's MapReduce
•     Spark vs MapReduce
•     Spark as an Alternative to Apache Tez
•     The Resilient Distributed Dataset (RDD)
•     Spark Streaming (Micro-batching)
•     Spark SQL
•     Example of Spark SQL
•     Spark Machine Learning Library
•     GraphX
•     Spark vs R

Chapter 3. Hadoop Distributed File System Overview

•     Hadoop Distributed File System (HDFS)
•     HDFS High Availability
•     HDFS "Fine Print"
•     Storing Raw Data in HDFS
•     Hadoop Security
•     HDFS Rack-awareness
•     Data Blocks
•     Data Block Replication Example
•     HDFS NameNode Directory Diagram
•     Accessing HDFS
•     Examples of HDFS Commands
•     Other Supported File Systems
•     WebHDFS
•     Examples of WebHDFS Calls
•     Client Interactions with HDFS for the Read Operation
•     Read Operation Sequence Diagram
•     Client Interactions with HDFS for the Write Operation
•     Communication inside HDFS

Chapter 4. The Spark Shell

•     The Spark Shell
•     The Spark Shell UI
•     Spark Shell Options
•     Getting Help
•     The Spark Context (sc) and SQL Context (sqlContext)
•     The Shell Spark Context
•     Loading Files
•     Saving Files
•     Basic Spark ETL Operations

Chapter 5. Spark RDDs

•     The Resilient Distributed Dataset (RDD)
•     Ways to Create an RDD
•     Custom RDDs
•     Supported Data Types
•     RDD Operations
•     RDDs are Immutable
•     Spark Actions
•     RDD Transformations
•     Other RDD Operations
•     Chaining RDD Operations
•     RDD Lineage
•     The Big Picture
•     What May Go Wrong
•     Checkpointing RDDs
•     Local Checkpointing
•     Parallelized Collections
•     More on parallelize() Method
•     The Pair RDD
•     Where do I use Pair RDDs?
•     Example of Creating a Pair RDD with Map
•     Example of Creating a Pair RDD with keyBy
•     Miscellaneous Pair RDD Operations
•     RDD Caching
•     RDD Persistence
•     The Tachyon Storage

Chapter 6. Shared Variables in Spark

•     Shared Variables in Spark
•     Broadcast Variables
•     Creating and Using Broadcast Variables
•     Example of Using Broadcast Variables
•     Accumulators
•     Creating and Using Accumulators
•     Example of Using Accumulators
•     Custom Accumulators

Chapter 7. Parallel Data Processing with Spark

•     Running Spark on a Cluster
•     Spark Stand-alone Option
•     The High-Level Execution Flow in Stand-alone Spark Cluster
•     Data Partitioning
•     Data Partitioning Diagram
•     Single Local File System RDD Partitioning
•     Multiple File RDD Partitioning
•     Special Cases for Small-sized Files
•     Parallel Data Processing of Partitions
•     Spark Application, Jobs, and Tasks
•     Stages and Shuffles
•     The "Big Picture"

Chapter 8. Introduction to Spark SQL

•     What is Spark SQL?
•     Uniform Data Access with Spark SQL
•     Hive Integration
•     Hive Interface
•     Integration with BI Tools
•     Spark SQL is No Longer Experimental Developer API!
•     What is a DataFrame?
•     The SQLContext Object
•     The SQLContext API
•     Changes Between Spark SQL 1.3 to 1.4
•     Example of Spark SQL (Scala Example)
•     Example of Working with a JSON File
•     Example of Working with a Parquet File
•     Using JDBC Sources
•     JDBC Connection Example
•     Performance & Scalability of Spark SQL

Chapter 9. Graph Processing with GraphX

•     What is GraphX?
•     Supported Languages
•     Vertices and Edges
•     Graph Terminology
•     Example of Property Graph
•     The GraphX API
•     The GraphX Views
•     The Triplet View
•     Graph Algorithms
•     Graphs and RDDs
•     Constructing Graphs
•     Graph Operators
•     Example of Using GraphX Operators
•     GraphX Performance Optimization
•     The PageRank Algorithm
•     GraphX Support for PageRank

Chapter 10. Machine Learning Algorithms

•     Supervised vs Unsupervised Machine Learning
•     Supervised Machine Learning Algorithms
•     Unsupervised Machine Learning Algorithms
•     Choose the Right Algorithm
•     Life-cycles of Machine Learning Development
•     Classifying with k-Nearest Neighbors (SL)
•     k-Nearest Neighbors Algorithm
•     k-Nearest Neighbors Algorithm
•     The Error Rate
•     Decision Trees (SL)
•     Random Forests
•     Unsupervised Learning Type: Clustering
•     K-Means Clustering (UL)
•     K-Means Clustering in a Nutshell
•     Regression Analysis
•     Logistic Regression

Chapter 11. The Spark Machine Learning Library

•     What is MLlib?
•     Supported Languages
•     MLlib Packages
•     Dense and Sparse Vectors
•     Labeled Point
•     Python Example of Using the LabeledPoint Class
•     LIBSVM format
•     An Example of a LIBSVM File
•     Loading LIBSVM Files
•     Local Matrices
•     Example of Creating Matrices in MLlib
•     Distributed Matrices
•     Example of Using a Distributed Matrix
•     Classification and Regression Algorithm
•     Clustering

Chapter 12. Spark Streaming

•     What is Spark Streaming?
•     Spark Streaming as Micro-batching
•     Use Cases
•     Some "Competition"
•     Spark Streaming Features
•     How It Works
•     Basic Data Stream Sources
•     Advanced Data Stream Sources
•     The DStream Object
•     DStream - RDD Diagram
•     The Operational DStream API
•     DStream Output Operations
•     The
•     StreamingContext Object
•     TCP Text Streams Example (in Scala)
•     Accessing the Underlying RDDs
•     The Sliding Window Concept
•     The Sliding Window Diagram
•     The Window Operations
•     A Windowed Computation Example (Scala)
•     Points to Remember
•     Other Points to Remember

Lab Exercises

    Lab 1. Learning the Lab Environment
    Lab 2. Elements of Functional Programming with Python
    Lab 3. The Hadoop Distributed File System
    Lab 4. Using the spark-submit Tool
    Lab 5. The Spark Shell
    Lab 6. RDD Performance Improvement Techniques
    Lab 7. Spark ETL and HDFS Interface
    Lab 8. Using Broadcast Variables
    Lab 9. Using Accumulators
    Lab 10. Common Map / Reduce Programs in Spark
    Lab 11. Spark SQL
    Lab 12. Getting Started with GraphX
    Lab 13. PageRank with GraphX
    Lab 14. Using Random Forests for Classification with Spark MLlib
    Lab 15. Using k-means Algorithm from MLlib
    Lab 16. Text Classification with Spark ML Pipeline
    Lab 17. Spark Streaming: Part 1
    Lab 18. Spark Streaming: Part 2

Learning Path

• Please contact us for suggestions.

• Machine Learning with Apache Spark

Ways to Attend

• Attend a public course, if there is one available. Please check our schedule, or register your interest in joining a course in your area.
• Private onsite Team training also available, please contact us to discuss. We can customise this course to suit your business requirements.