Developing Hadoop Applications

Winter 2015

Developing Hadoop Applications

This Lab Guide is protected under U.S. and international copyright laws, and is the exclusive property of
MapR Technologies, Inc. 2015, MapR Technologies, Inc. All rights reserved.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

ii

Get Started
Welcome to Developing Hadoop Applications
This course teaches developers, via lectures and hands-on lab exercises, how to write Hadoop
Applications using MapReduce in Java. This course prepares you to successfully pass the MapR Certified
Hadoop Professional Developer, provided that you complete all lab exercises. Covered are:

Describe the MapReduce programming paradigm

Use MRv1 and MRv2 job frameworks

Implement a basic MapReduce application

Use the MapReduce API

Manage and test MapReduce jobs

Tune MapReduce performance

Launch and manage multiple jobs

Work with different data in MapReduce

Implement and manage streaming MapReduce jobs

Prerequisites

Beginner-to-intermediate fluency with Java

A laptop (Linux, PC or Mac) to connect to a Hadoop cluster via SSH and web browser

Basic Linux end-user skills (e.g ls, cd, vi)

How to Use this Guide

Typographical Conventions
Courier

Things you type as shown

Courier

Things as seen on screen

italics

Things you replace with a value

Durations

Suggested time it takes to complete an exercise. Some are required, other are optional

Developing Hadoop Applications

Lab Exercise Timing

Each Lessons Labs are listed with estimated time to completion.
Exercise
Connect to the MapR cluster (sandbox or AWS)

Required

Duration

Yes

30 min

Tools Needed to Complete Labs

Please download all associated course materials. You will find documents that offer instructions on
setting up the lab environment you will need to complete the labs:

MapR Hadoop Training Sandbox Download the sandbox from here. You will need to download
and install VirtualBox to use this sandbox..

Amazon Web Services Node instructions on how you can purchase and set up your own lab
environment in the cloud.

doc.mapr.com a repository of documentation on MapReduce

Connect to the MapR Cluster (Sandbox or AWS)

The purpose of this lab is for you to establish connectivity to your MapR cluster environment that you
will use for the hands-on exercises in this course.

Logging into the Sandbox

This procedure assumes that you have downloaded and installed the MapR Sandbox as part of the class
prerequisites. If not, please do that before continuing here. Make a note of the IP address assigned to
the Sandbox, as you will need it to log in remotely.
The Sandbox is preconfigured with several user accounts. You will be logging in as user01 to complete
most of the lab exercises.
You will need to use a Secure Shell (SSH) client to log into the Sandbox. PuTTY, Cygwin, or a bash shell
on a Unix-like OS with OpenSSH installed are all acceptable clients. If you do not have an SSH client, you
can log in to the Sandbox using the virtual machine console mode. Connect to your Sandbox as user01.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

OPTION 1: Log in using PuTTY

1. Start your PuTTY Client.

2. In the Host Name (or IP address) box, enter the host name or IP address for your Sandbox virtual
machine.
3. Click the Open button to open the connection to your node.
4. When the terminal window appears, log into the terminal as user01.
OPTION 2: Log in using terminal SSH client
Start your terminal client. Use the ssh program to connect to the hostname or IP address for your
Sandbox. Connect as user01 , password mapr
$ ssh user01@IP-address

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

OPTION 3: Log in using Virtual Machine Console

You can log in directly to the Sandbox console (in VMWare Player or Workstation, or in Virtual Box).
=== MapR Sandbox for Hadoop ===
Version: 1.0
MapR Sandbox for Hadoop Installation finished successfully.
Please go to http://192.168.30.129:8443 to begin your experience
Open a browser on your host machine
And enter the URL in the browser address field
Log in to this virtual machine: Linux/Windows <Alt + F2>, Mac OS
X <Option + F5>
Use the appropriate key combination for your host machine (Alt + F2 for Windows or Linux, Option + F5
for Mac OS X) to reach the login prompt.
CentOS release 6.5 (Final)
Kernel 2.6.32-431.e16.x86_64 on an x86_64
Maprdemo login: _

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lesson 1: Introduction to Developing

Hadoop Applications
In the lessons lab exercises, you will run a few MapReduce jobs from the command line and examine
job information in the MapR Console Server (MCS).
IMPORTANT NOTE: Some of the commands in this lab are too long to fit on a single line in this
document. The backslash character in a command indicates that the command continues on the next
line. You don't need to type the backslash character at all, but if you do, make sure the very next
character following the backslash is a carriage return.
Exercise

Duration

1.1: Run wordcount - Explore running the classic "word count"

MapReduce application that ships with Hadoop.

20 min

1.2: Examine Job Metrics in MCS - Use the MCS to examine metrics for
jobs you previously launched.

10 min

Exercise 1.1: Run wordcount

Run wordcount Against a Text File
1. Create a directory in your home directory as follows:
$ mkdir /user/user01/1.1

2. Create a file in that directory as follows.

$ echo "Hello world! Hello" > /user/user01/1.1/in.txt

3. Run the MRv1 version of the wordcount application against the input file.
$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/hadoop-0.20.2-dev-\
examples.jar wordcount /user/user01/1.1/in.txt /user/user01/1.1/OUT

4. Check the output of the wordcount application.

$ cat /user/user01/1.1/OUT/part-r-00000

Developing Hadoop Applications

Run wordcount Against a Set of Text Files

5. Create a directory in your home directory as follows:
$ mkdir -p /user/user01/1.1/IN2

6. Create a set of text files in that directory as follows.

$ cp /etc/*.conf /user/user01/1.1/IN2 2>/dev/null

7. Determine how many files are in that directory.

$ ls /user/user01/1.1/IN2 | wc -l

8. Run the MRv2 version of the wordcount application against the directory.
$ hadoop2 jar /opt/mapr/hadoop/hadoop-\
2.4.1/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.4.1-mapr-\
1408.jar wordcount /user/user01/1.1/IN2 /user/user01/1.1/OUT2

9. Check the output of the wordcount application.

$ wc -l /user/user01/1.1/OUT2/part-r-00000
$ more /user/user01/1.1/OUT2/part-r-00000

Run wordcount Against a Binary File

1. Create a directory in your home directory as follows:
$ mkdir -p /user/user01/1.1/IN3

2. Create a binary file in that directory as follows.

$ cp /bin/cp

/user/user01/1.1/IN3/mybinary

3. Verify the file is a binary.

$ file

/user/user01/1.1/IN3/mybinary

4. See if there is any readable text in the binary.

$ strings

/user/user01/1.1/IN3/mybinary | more

5. Run the MRv1 version of the wordcount application (using the MRv2 client) against the input
file. This will show that binaries compiled for MRv1 will run in a MRv2 framework.
$ hadoop2 jar /opt/mapr/hadoop/hadoop-0.20.2/hadoop-0.20.2-dev-\
examples.jar wordcount /user/user01/1.1/IN3/mybinary \
/user/user01/1.1/OUT3

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

6. Check the output of the wordcount application.

$ more /user/user01/1.1/OUT3/part-r-00000

7. Cross-reference the frequency of the word ATUH in the binary and in the wordcount output.
$ strings /user/user01/1.1/IN3/mybinary | grep -c ATUH
$ egrep -ac ATUH /user/user01/1.1/OUT3/part-r-00000

Exercise 1.2: Examine Job Metrics in MCS

1. Connect to the MCS in your Web browser.
https://node-IP:8443

2. Login as the 'mapr 'user with password 'mapr'.

3. Navigate to Navigation System Settings Metrics
4. In the "Configure Metrics Database" dialog, fill in the form as follows and click the "OK" button:
a. URL = localhost:3306
b. Username = mapr
c. Password = mapr
d. Schema = metrics
5. Scroll in the listing at the bottom of the view to find your jobs (as a combination of the Job
Name and User fields).
6. Click the link word count for one of the jobs you launched.

How many tasks comprised that job?

How long did they each last?

On which node did they run? Note that in a single-node cluster, there's only one
machine the job can run on.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lesson 2: Job Execution Framework

MapReduce v1 & v2
In the lessons lab exercises, you will run a YARN job from the command line and examine job
information in the History Server Web UI.

Exercise

Duration

2.1: Run distributed shell - Explore running the sample distributed

shell YARN application that ships with Hadoop.

10 min

2.2: Examine Job Results in History Server - Use the History Server
Web UI to examine metrics for jobs you previously launched.

10 min

Exercise 2.1: Run DistributedShell

In this exercise, you will explore running a sample YARN job that ships with Hadoop.

Run DistributedShell With a Shell Command

1. Launch the YARN job using the yarn command.
$

yarn jar /opt/mapr/hadoop/hadoop-2.4.1/share/hadoop/yarn/hadoop-yarn-\

applications-distributedshell-2.4.1-mapr-1408.jar -shell_command \
/bin/ls -shell_args /user/user01 -jar /opt/mapr/hadoop/hadoop-\
2.4.1/share/hadoop/yarn/hadoop-yarn-applications-distributedshell-\
2.4.1-mapr-1408.jar

2. When the job completes, scroll back through the output to determine your container ID for the
shell. HNT look at the example output below.

Developing Hadoop Applications

15/01/21 18:34:07 INFO distributedshell.Client: Got application

report from ASM for, appId=1, clientToAMToken=null,
appDiagnostics=, appMasterHost=yarn-training/192.168.56.102,
appQueue=root.user01, appMasterRpcPort=-1,
appStartTime=1421894036331, yarnAppState=FINISHED,
distributedFinalState=SUCCEEDED, appTrackingUrl=http://yarntraining:8088/proxy/application_1421893926516_0001/A,
appUser=user01

Check Standard Output and Standard Error for Job

3. Change directory to the output directory for YARN jobs.
$ cd /opt/mapr/hadoop/hadoop-2.4.1/logs/userlogs

4. List the contents of the directory.

$ ls
5. Change directory to your application output directory.
$ cd application_timestamp_appid
6. List the contents of the directory.
$ ls
7. Change directory to the second container output directory.
$ cd container_timestamp_appid_01_000002

8. Display the contents of the stdout file. You should see a listing of the /user/user01
directory.
$ cat stdout
9. Display the contents of the stderr file. It should be empty.
$ cat stderr

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 2.2: Examine Results in History Server Web UI

In this exercise, you will use the Web UI provided by the History Server to examine information for the
job you previously launched.
1. Connect to the History Server in your Web browser.
http://node-IP:8088

2. Scroll through the applications to find your application ID.

3. Click the link associated with your YARN job.
4. How long did the job take?

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lesson 3: Modify a MapReduce

Program
Lab Overview
The labs for this lesson cover how to make some modifications to an existing MapReduce program,
compile it, run it, and examine the output. The goal of these lab exercises is to calculate the mean
surplus or deficit. The existing code calculates the minimum and maximum values. Your output will
include these values with the year that those extrema occurred.
Here is the output expected from the map and reduce code for this lab:
min(2009):

-1412688.0

max(2000):

236241.0

mean:

-93862.0

Exercise
3.1: Copy lab files into your home directory

Duration
5 min

3.2: Modify the driver

5 min

3.3: Compile and run the map-only program

5 min

3.4: Implement code in the reducer

15 min

3.5: Compile and run the MapReduce program

10 min

Summary of Data
The data set were using is the history of the United States federal budget from the year 1901 to 2012.
The data was downloaded from http://www.whitehouse.gov/omb/budget/Overview and has been
massaged for this exercise.
Here is a sample record from the data set:
1968 152973 178134 -25161 128056 155798 -27742 24917 22336 2581

Developing Hadoop Applications

The fields of interest in this exercise are the first and fourth fields (year and surplus or deficit). The
second field is the total income derived from federal income taxes, and the third field is the
expenditures for that year. The fourth field is the difference between the second and third fields. A
negative value in the fourth field indicates a budget deficit and a positive value indicates a budget
surplus.

Exercise 3.1: Copy the Lab Files

1. Create a directory as follows:
$ mkdir /user/user01/3

2. Copy the source code from your laptop into that directory
$ scp RECEIPTS_LAB.zip user01@node-ip:/user/user01/3

3. Change directory into the directory.

$ cd /user/user01/3
4. Unzip the RECEIPTS_LAB.zip file.
$ unzip RECEIPTS_LAB.zip

Exercise 3.2: Modify Code in the Driver

1. Change directory into the RECEIPTS_LAB directory.
$ cd /user/user01/3/RECEIPTS_LAB

2. Open the ReceiptsDriver.java source file with your favorite text editor.
$ vi ReceiptsDriver.java

3. Make the following changes to the driver

//TODO Change the name of the job to a string of your choice
HINT: Change the string my receipts to any string of your choice
//TODO Comment out the Reducer definition
HINT: The reducer is defined by the setReducerClass() method of the Job object
4. Save the ReceiptsDriver.java file.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 3.3: Compile and Run the Map-only MapReduce

Program
1. Change directory into the RECEIPTS_LAB directory.
$ cd /user/user01/3/RECEIPTS_LAB

2. Execute the rebuild.sh script to compile your code.

$ ./rebuild.sh

3. Execute the rerun.sh script to run your code. This script will execute your code.
$ ./rerun.sh

4. Examine the output from your MapReduce job. Note you may need to wait a minute before the
job output is completely written to the output files.
$ cat /user/user01/3/RECEIPTS_LAB/OUT/part*

Here is partial output expected for this exercise:

summary

1901_63

summary

1902_77

summary

1903_45

summary

1904_-43

summary

1905_-23

summary

1906_25

summary

1907_87

summary

1908_-57

summary

1909_-89

summary

1910_-18

summary

1911_11

If you did not obtain the results above, youll need to revisit your Mapper class. Ask your instructor for
help if you need. Once you obtain the correct intermediate results from the map-only code, proceed to
the next exercise.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 3.4: Implement Code in the Reducer

In this exercise, you will implement code in the reducer to calculate the mean value. The code has
already been provided to calculate minimum and maximum values.
Recall that your mapper code will produce intermediate results. One such record looks like this:
summary 1968_-25161

When you execute the code for this lab, there will only be one reducer (since there is only one key
summary). That reducer will iterate over all the intermediate results and pull out the year and surplus
or deficit. Your reducer will keep track of the minimum and maximum values (as temp variables) as well
as the year those values occurred. You will also need to keep track of the sum of the surplus or deficit
and count of the records in order to calculate the mean value.
Recall that the output of your reducer will look like this:
min(2009):

-1412688.0

max(2000):

236241.0

mean: -93862.0

1. Change directory into the RECEIPTS_LAB directory.

$ cd /user/user01/3/RECEIPTS_LAB

2. Open the ReceiptsReducer.java source file with your favorite text editor.
$ vi ReceiptsReducer.java

3. Make the following changes to the reducer

//TODO declare and initialize int values for sum and count
//TODO increment the sum by tempValue
//TODO increment the count by 1
//TODO declare and calculate float value for mean
//TODO set the keyText to "mean:"
//TODO write keyText and the mean to the context

4. Save the ReceiptsReducer.java file.

5. Open the ReceiptsDriver.java source file with your favorite text editor and make the
following changes :
//TODO uncomment the Reducer definition

6. Save the ReceiptsDriver.java file.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 3.5: Compile and Run Your Code

1. Change directory into the RECEIPTS_LAB directory.
$ cd /user/user01/3/RECEIPTS_LAB

2. Execute the rebuild.sh script to compile your code.

$ ./rebuild.sh

3. Execute the rerun.sh script to run your code.

$ ./rerun.sh

4. Examine the output from your MapReduce job.

$ cat /user/user01/3/RECEIPTS_LAB/OUT/part*

Here is the output expected for this exercise:

min(2009):

-1412688.0

max(2000):

236241.0

mean: -93862.0

Note: The solution for this lab is found in RECEIPTS_SOLUTION.zip.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lesson 4: Write a MapReduce Program

Lab Overview
The high level objective of this lab is to write your first complete MapReduce program using the
numerical summary pattern weve been focusing on. The lab provides generic templates for the map,
reduce, and driver classes. The exercises guide you through how to calculate the minimum, maximum,
and mean SAT verbal and SAT math scores over the whole data set.
You should use the lab solution for Lab 4 to guide you through this lab. The overall structure of both
solutions is the same since it follows the same min-max-mean pattern. How you parse records in this
lab will be different because the format of the UNIVERSITY records is different than the format of the
RECEIPTS records.
To get started, login to a MapR node as your user and perform the following exercises.
Exercise
4.1: Copy lab files into your directory

Duration
5 min

4.2: Implement the mapper class

20 min

4.3: Implement the mapper class

45 min

4.4: Implement the reducer class

4.5: Compile and run the MapReduce program

10 min
10 min

Developing Hadoop Applications

Summary of Data
This lab examines data sampled from university students across North America. The data set can be
downloaded from http://archive.ics.uci.edu/ml/datasets/University.
Not every record contains the same number of fields, but every record starts with the string (def-
instance and ends with the string )). Each record contains information for a single university in the
survey. Here is a sample record:
(def-instance Adelphi
(state newyork)
(control private)
(no-of-students thous:5-10)
(male:female ratio:30:70)
(student:faculty ratio:15:1)
(sat math 475)
(expenses thous$:7-10)
(percent-financial-aid 60)
(no-applicants thous:4-7)
(percent-admittance 70)
(percent-enrolled 40)
(academics scale:1-5 2)
(social scale:1-5 2)
(sat verbal 500)
(quality-of-life scale:1-5 2)
(academic-emphasis business-administration)
(academic-emphasis biology))

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 4.1: Copy the Lab Files to Your Directory

1. Create a directory in your directory as follows:
$ mkdir /user/user01/4

2. Copy the lab ZIP file into that directory from your laptop
$ scp UNIVERSITY_LAB.zip user01@node-IP:/user/user01/4

3. Change directory into your directory.

$ cd /user/user01/4

1. Unzip the lab ZIP file.

$ unzip UNIVERSITY_LAB.zip

Exercise 4.2: Implement the Driver Class

1. Change directory into the UNIVERSITY_LAB directory.
$ cd /user/user01/4/UNIVERSITY_LAB

2. Open the UniversityDriver.java source file with your favorite text editor.
$ vi UniversityDriver.java

3. Implement each TODO in the UniversityDriver.java file as follows:

// TODO check the CLI
// TODO: instantiate a new job with a unique name
// TODO: define driver, mapper, and reducer classes in job
// TODO: define input format for job
// TODO: define record delimiter in input format class to "))"

HINT: conf.set("textinputformat.record.delimiter","))");
// TODO define output key and output value class
// TODO define input and output paths
// TODO: launch job synchronously
// TODO: instantiate a new configuration object
// TODO: call the ToolRunner run() method

4. Save the UniversityDriver.java file.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 4.3: Implement the Mapper Class

Recall that a single record contains an unknown number of fields after the start of the record and before
either the sat math or sat verbal field.
(def-instance xxx
. . .
(sat verbal 500)
. . .
(sat math 475)
. . .))

Or the sat math field may come before the sat verbal field. Or one or both of the sat math and
sat verbal fields may not be part of the record at all. Skip records that do not contain both fields.
NOTE: the data set is not uniform from beginning to end. Examine the first few records in the file, and
then skip to line 1000 or so in the file. Use the less or view command to examine the input data
set (UNIVERSITY_LAB/DATA/university.txt).
1. Change directory into the UNIVERSITY_LAB directory.
$ cd /user/user01/4/UNIVERSITY_LAB

2. Open the UniversityMapper.java source file with your favorite text editor.
$ vi UniversityMapper.java

3. Implement each TODO in the UniversityMapper.java file as follows:

// TODO convert record (value) to String
// TODO determine if the record contains fields for both "sat verbal"
AND "sat math"
// TODO iterate through tokens until you find "sat verbal" field
// TODO split the "sat verbal" field token
// TODO pull out the score from the "sat verbal" field
// TODO continue through iterator until you find "sat math" field
// TODO split the "sat math" field token
// TODO pull out the score from the "sat math" field
// TODO emit key-value as

"satv", score

// TODO emit key-value as "satm", score

4. Save the UniversityMapper.java file.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 4.4: Implement the Reducer Class

Recall that one reducer will be given a list of key-value pairs that looks like this:
satv 480 500 530 . . .

The other reducer will be given a list of key-value pairs that looks like this:
satm 400 500 510 . . .

1. Change directory into the UNIVERSITY_LAB directory.

$ cd /user/user01/4/UNIVERSITY_LAB

2. Open the UniversityReducer.java source file with your favorite text editor.
$ vi UniversityReducer.java

3. Implement each TODO in the UniversityReducer.java file as follows:

// TODO loop through the values and calculate min, max, and mean
// TODO emit keystring_min, min
// TODO emit keystring_max, max
// TODO emit keystring_mean, mean

4. Save your UniversityReducer.java file.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 4.5: Compile and Run the MapReduce Program

Compile the MapReduce program
1. Launch the rebuild.sh script to recompile the source code.
$ ./rebuild.sh

2. If you have any errors you cant resolve, you may wish to check the output from your map phase
by setting mapred.num.reduce.tasks to 0 in your configuration.

Run the MapReduce Program

Launch the rerun.sh script to execute the code.
$ ./rerun.sh

Check the Output of the Program

$ cat /user/user01/4/UNIVERSITY_LAB/OUT/part-r-00000

NOTE: the solution for this exercise is located in UNIVERSITY_SOLUTION.zip.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lesson 5: Manage, Monitor, and Test

MapReduce Jobs
The goal of the exercises in the lessons lab is to use the tools and methods described in the lecture to
manage and debug your MapReduce jobs and applications.
Login to one of the nodes in the MapR cluster as your user and perform the following exercises. Note
that some commands may have multiple lines separated by a backslash character (\). This indicates
that the command should continue without typing a carriage return.

Exercise

Duration

5.1: Copy lab files to your directory

5 min

5.2: Examine default job output

30 min

5.3: Write MapReduce code to use custom counters

30 min

5.4: Write MapReduce code to use standard output, error, and logging

30 min

5.5: Use MRUnit to test a MapReduce application

45 min

5.6: Use the hadoop CLI to manage jobs

30 min

Developing Hadoop Applications

Exercise 5.1: Copy the Lab Files into Your Directory

1. Create a directory in your directory as follows:
$ mkdir /user/user01/5

2. Copy the source code into that directory

$ scp SLOW_LAB.zip user01@node-IP:/user/user01/5
$ scp VOTER_LAB.zip user01@node-IP:/user/user01/5

3. Change directory into the directory.

$ cd /user/user01/5

4. Unzip the source code ZIP files.

$ unzip SLOW_LAB.zip
$ unzip VOTER_LAB.zip

5. Uncompress the data file for the VOTER_LAB

$ gunzip /user/user01/5/VOTER_LAB/DATA/myvoter.csv.gz

6. Inject some faulty records into your data set. For example:
$ echo "0,james,14,independent,100,100" >>\
/user/user01/5/VOTER_LAB/DATA/myvoter.csv
$ echo "0,james,25" >>\
/user/user01/5/VOTER_LAB/DATA/myvoter.csv

Exercise 5.2: Examine Default Job Output

In this exercise, you will run the teragen and terasort MapReduce applications from the examples
provided in the Hadoop distribution. You will then examine the records produced from running each
one.

Run teragen
1. Run the teragen MapReduce application to generate 1000 records.
$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/\
hadoop-0.20.2-dev-examples.jar \
teragen 1000 /user/user01/5/TERA_IN

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

2. Look at the teragen job output counters to examine output

Question: Why are there no input or output records for the reducer in the job output?
Answer: The teragen application is a map-only application.
2. Examine the files produced by teragen and answer the questions below.

What type of file is produced?

$ file /user/user01/5/TERA_IN/part-m-0000*

Why is the number of records we generated with teragen different than the total number
of lines in the files?
$ wc -l /user/user01/5/TERA_IN/part-m-0000*

Can you make sense out of the files by looking at them?

$ view /user/user01/5/TERA_IN/part-m-00000

Run terasort
1. Run the terasort application to sort those records you just created and look at the job
output.
$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/\
hadoop-0.20.2-dev-examples.jar \
terasort /user/user01/5/TERA_IN /user/user01/5/TERA_OUT

$ hadoop1 job -history /user/user01/5/TERA_OUT

2. (Records) Look at the terasort standard output to determine the following:

Number of mappers launched

Question: is this equal to the number of input files?
Answer: yes (2)

Number of map and reduce input and output records.

Question: When would the number of map input records be different than the number of
map output records?
Answer: if the map method is doing any sort of filtering (e.g. dropping "bad" records).

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Number of combine input and output records

Question: what does this imply about the terasort application?
Answer: terasort does not use a combiner.

Exercise 5.3: Write MapReduce Code with Custom Job

Counters
In this exercise, you will write the logic to identify a bad record in a data set, then define a custom
counter to count bad records from that data set. This is what a good record looks like:
1,david davidson,10,socialist,369.78,5108

There are 6 fields total a primary key, name, age, party affiliation, and two more fields you dont care
about. You will implement a record checker that validates the following:

there are exactly 6 fields in the record

third field is a reasonable age for a voter

1. Change directory to the VOTER_LAB directory.

$ cd /user/user01/5/VOTER_LAB

2. Open the VoterDriver.java file with the view command. What character separates the
keys from the values in the records? Close the file.
3. Open the VoterMapper.java file with your favorite editor. Which character is the value of
the record tokenizing on? Keep the file open for the next step.
4. Implement the TODOs below to validate the record.
//TODO check number of tokens in iterator
//TODO increment "wrong number of fields" counter and return if wrong
//TODO eat up the first two tokens
//TODO convert age to an int
//TODO validate the age is a reasonable age
//TODO increment "bad age" counter and return if bad
//TODO convert age to an IntWritable
//TODO pull out party

from record

//TODO emit key-value as party-age

5. Compile and execute the code. Based on the minimum, maximum, and mean values for voter
ages, what do you conclude about the nature of the data set?

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

6. Examine the output in your terminal from the job to determine the number of bad records.

How many records have the wrong number of fields?

How many records have a bad age field?

Does the total number of bad records plus the total number of reduce input records equal
the total number of map input records?

Exercise 5.4: Write MapReduce Code to Use Standard

Output, Error, and Logging
In this exercise, you will generate standard error and log messages and them consume them in the MCS.

Modify the mapper to send messages to standard error

In this task, use standard error instead of job counters to keep track of bad records.
1. Change directory to the VOTER_LAB directory.
$ cd /user/user01/5/VOTER_LAB

2. Open the VoterMapper.java file with your favorite editor.

$ vi VoterMapper.java

3. Implement the following TODOs in the code

Instead of incrementing the bad record counter for incorrect number of tokens, write a
message to standard error. Include the bad record in the message.
HINT: Use System.err.println()

Instead of incrementing the bad record counter for invalid age, write a message to standard
error. Include the bad record in the message.
HINT: Use System.err.println()

4. Save the file.

5. Compile and execute the application.
$ ./rebuild.sh
$ ./rerun.sh "-DXmx1024m"

6. Examine standard error messages for your job in MCS as follows:

Login to the MCS

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Click the Cluster Jobs link in the navigation pane

Scroll through the jobs in the content pane to find the job you just launched. Look for the
name you specified when you ran the job.

Click the link to the job name.

Click the link to the primary attempt of the map task.

Click the Log>> link in the Log Location field of the task attempt.

Modify the mapper to write messages to syslog

In this task, you will perform the same logic as in the previous task, except you'll write the message to
syslog.
1. Change directory to the VOTER_LAB directory.
$ cd /user/user01/5/VOTER_LAB

2. Open the VoterMapper.java file with your favorite editor.

$ vi VoterMapper.java

3. Implement the following TODOs in the code

Instead of incrementing the bad record counter or writing to standard error for incorrect
number of tokens, write a message to syslog. Include the bad record in the message.
HINT: Use log.error() from Apache Commons Logging

Instead of incrementing the bad record counter or writing to standard error for invalid age,
write a message to syslog. Include the bad record in the message.
HINT: Use log.error() from Apache Commons Logging

4. Save the file.

5. Compile and execute the application. Replace the username variable with your login user
name.
$ ./rebuild.sh
$ ./rerun.sh "-DXmx1024m"

Examine syslog messages for your job in MCS

In this task, you will use the MCS to consume syslog messages.
1. Login to the MCS

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

2. Click the Cluster Jobs link in the navigation pane

3. Scroll through the jobs in the content pane to find the job you just launched. Look for the name
you specified when you ran the job
4. Click the link to the job name
5. Click the link to the primary attempt of the map task.
6. Click the Log>> link in the Log Location field of the task attempt. (you may have to specify
the IP address)

Exercise 5.5: Use MRUnit to Test a MapReduce Application

In this exercise, the code to test the mapper using MRUnit is already provided. You will follow that
example to implement the reducer test.
Recall the VoterMapper map method emits the key-value pair: (party, age). For example, with input
"1,david davidson,20,socialist,369.78,5108" you should expect output (socialist, 20).

Test the mapper against different input

1. Change directory to the VOTER_LAB directory.
$ cd /user/user01/5/VOTER_LAB

2. View the mymaptest.dat file with the cat command. The first line of the file is the input
record. The second line of the file is the output from the map method for the input.
$ cat mymaptest.dat

3. Test the map method against the test file you should get a "success" message.
$ ./retest.sh map mymaptest.dat

4. Now edit the test file so that the input and expected output do not match.
5. Test the map method against the test file this time you should get an exception
$ ./retest.sh map mymaptest.dat

Implement code to test the reducer

1. Change directory to the VOTER_LAB directory.
$ cd /user/user01/5/VOTER_LAB

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

2. View the myreducetest.dat file with the cat command. The first line of the file is the
input record (including the key and list of values). The second line of the file is the output from
the reduce method for the associated input.
$ cat myreducetest.dat

3. Implement the TODO in the VoterTest.java file to write the unit test for the reducer:
//TODO declare the reduceDriver object
//TODO instantiate a reducer object and reducer driver
//TODO implement the testReducer method
//TODO create a list object for reduce input
//TODO tokenize the first line from the input file
//TODO pull out the key from the tokenized line
//TODO loop through tokens to add IntWritable to reduce input list

Build and execute the unit test for the reducer

1. Change directory to the VOTER_LAB directory.
$ cd /user/user01/5/VOTER_LAB

2. Edit the VoterReducer.java file such that only the key-value pair for the mean is emitted
(comment out the lines that write the min and max to the context). This is because we are only
examining the output for mean in our MRUnit test.
$ vi VoterReducer.java
3. Rebuild the jar file.
$ ./rebuild.sh

4. Run the reduce test against the input file.

$ ./retest.sh reduce myreducetest.dat

5. Make a change in the myreducetest.dat file so that the expected output intentionally does not
match the expected output. Then retest the reduce method to see what the error looks like.
$ ./retest.sh reduce myreducetest.dat

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 5.6: Use the Hadoop CLI to Manage Jobs

In this exercise, we use the Hadoop CLI to manage jobs.

Launch a long-running job

1. Change directory to the SLOW_LAB directory.
$ cd /user/user01/5/SLOW_LAB

2. Launch the MapReduce application and specify the sleep time (in ms).
$ ./rerun.sh "-DXmx1024m -D my.map.sleep=4000"

Get counter values for job

1. Log in to your cluster in a second terminal window.
2. Find the job id for your job
$ hadoop1 job -list

Display the job counter for MAPRFS_BYTES_READ. Replace the jobid variable using the output
from the previous command. NOTE: Wait till you see that the Job has started, before running this
command.
$ hadoop1 job -counter jobid FileSystemCounters MAPRFS_BYTES_READ

Kill the job

1. Kill your job. Replace the jobid variable appropriately.
$ hadoop1 job -kill jobid

2. Verify your job is no longer running.

$ hadoop1 job -list

Display job history

You cant show history for a failed job using the hadoop command, but you can examine the history
file. Use the hadoop command to display the history of the terasort application.
$ hadoop1 job history /user/user01/5/OUT

(Appendix: if the job hangs, change the memory allocation for filesystem in the warden.conf file.
service.command.mfs.heapsize.max=1024

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lab 6: Manage MapReduce Job

Performance
The goal of the exercises in this lessons lab is to use the tools and methods in lecture to learn to de-
tune a job and measure the performance.
Exercise

Duration

6.1: De-tune a job and measure its impact on performance

30 min

Exercise 6.1: De-Tune a Job and Measure Performance

Impact
In this exercise, you will actually "de-tune" a job and monitor its performance.

Establish a baseline for CPU time

In this task, you will run the teragen and terasort MapReduce applications from the examples provided
in the Hadoop distribution.
1. Create the lab directory.
$ mkdir /user/user01/6

2. Run the teragen MapReduce application to generate 1,000,000 records.

$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/hadoop-0.20.2-dev-\
examples.jar teragen 1000000 /user/user01/6/TERA_IN

3. Run the terasort application to sort those records you just created.
$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/hadoop-0.20.2-dev-\
examples.jar terasort DXmx1024m Dmapred.reduce.tasks=2 \
/user/user01/6/TERA_IN /user/user01/6/TERA_OUT_1

4. Determine the aggregate map phase run time of the job.

$ hadoop1 job -history /user/user01/6/TERA_OUT_1

5. Run it a few times more to establish a "good" baseline. Remove the output directory
/user/user01/6/TERA_OUT_1 before each run . Here are some values for a few runs. Fill
in the aggregate map phase run times below the ones given for your runs.

Developing Hadoop Applications

Run 1

Run 2

Run 3

Run 4

2m1s=121s

2m30s=150s

2m10s=130s

1m55s=115s

Modify the configuration and determine impact

In this task, you will run the teragen and terasort MapReduce applications from the examples
provided in the Hadoop distribution.
1. Run the terasort application to sort those records again, but this time with a modification in
the job parameters.
$ hadoop1 jar /opt/mapr/hadoop/hadoop-0.20.2/hadoop-0.20.2-dev-\
examples.jar terasort DXmx1024m Dmapred.reduce.tasks=2 \
-Dio.sort.mb=1 /user/user01/6/TERA_IN /user/user01/6/TERA_OUT_2

2. Determine the aggregate time spent in the map phase.

$ hadoop1 job -history /user/user01/6/TERA_OUT_2

3. Run it a few times more to establish a "good" test. Change the name of the output directory for
each rerun. Here are some values for a few runs. Fill in the results below the ones given for
your runs.

Run 1

Run 2

Run 3

Run 4

7m13s=433s

5m5s=305s

5m49s=349s

5m21s=321s

NOTE: there is a significant difference in time between the first run and the rest of the runs. Without
accounting for it, this is one reason we take several samples when benchmarking. Statistically, we
should probably throw out this outlier.
4. It appears that the change has impacted the amount of time spent in the map phase (which
makes sense given we are changing the io.sort.mb parameter). Calculate the speedup (or in
this case, slowdown) due to the configuration change.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Average aggregate time baseline = ( 121 + 150 + 130 + 115 ) / 4

= 129 seconds
Average aggregate time enhanced = ( 305 + 349 + 321 ) / 3
= 325 seconds
Speedup = baseline/enhanced - 1 = ( 129 / 325) - 1 = -0.60

In other words, the "enhanced" job performs 60% slower than the baseline.

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Lab 7: Working With Data

The goal of this exercise is to use the tools and methods learned in lecture to have a hands-on
experience of using other data. You will run a MapReduce program that uses HBase as a source.
Exercise
7.1: Run a MapReduce program to use HBase as source

Duration
30 min

Developing Hadoop Applications

Exercise 7.1: Run a MapReduce Program to Use HBase as

Source
In this exercise, you will create and populate a table in HBase to store the voter data from previous
exercises. You will then run a MapReduce program to calculate the usual maximum, minimum, and
mean values using data read from that table.

Create an HBase Table Using importtsv

In this task, you will use the importtsv utility to create an HBase table using the tab-separated
VOTER data we used in a previous lab. Log in to one of your cluster nodes to perform the steps in this
task an use the hbase command to create an empty table in MapR-FS.
1. Create a directory for this lab.
$ mkdir /user/uaer01/7

2. Copy the jar file to the lab directory.

$ scp VOTERHBASE_SOLUTION.zip user01@node-IP:/user/user01/7

3. Extract the jar file.

$ cd /user/user01/7
$ unzip VOTERHBASE_SOLUTION.zip

4. Launch the hbase shell:

$ hbase shell

5. Create the table in MapR-FS. Replace jcasaletto with your user name.
hbase> create '/user/user01/7/myvoter_table', {NAME => 'cf1'}, {NAME =>
'cf2'}, {NAME => 'cf3'}

6. Verify the file was created in MapR-FS.

hbase> quit
$ ls -l /user/user01/7/myvoter_table

7. Use the importtsv utility to import the data into the HBase table.
$ hadoop jar /opt/mapr/hbase/hbase-0.98.4/lib/hbase-server-\
0.98.4-mapr-1408.jar importtsv -Dimporttsv.columns=\
HBASE_ROW_KEY,cf1:name,cf2:age,cf2:party,cf3:contribution_amount,\
cf3:voter_number /user/user01/7/myvoter_table \
/user/user01/7/VOTERHBASE_SOLUTION/myvoter.tsv

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

8. Use the hbase command to validate the contents of the new table.
$ echo "scan '/user/user01/7/myvoter_table'" | hbase shell
ROW

COLUMN+CELL

column=cf1:name, timestamp=1406142938710, value=d

avid davidson

column=cf2:age, timestamp=1406142938710, value=49

column=cf2:party, timestamp=1406142938710, value=

socialist

column=cf3:contribution_amount, timestamp=1406142
938710, value=369.78

column=cf3:voter_number, timestamp=1406142938710,
value=5108

10

column=cf1:name, timestamp=1406142938710, value=o

scar xylophone

. . . <output omitted>
1000000 row(s) in 1113.9850 seconds

Run the MapReduce Program to Calculate Statistics

In this task, you will run your MapReduce program and then analyze the results.
1. Change directory to the location of the MapReduce program jar file.
$ cd /user/user01/7/VOTERHBASE_SOLUTION

2. Run the MapReduce program.

$ java -cp `hbase classpath`:VoterHbase.jar \
VoterHbase.VoterHbaseDriver \
/user/user01/7/myvoter_table \
/user/user01/7/OUT

3. Analyze the results (min, max, and mean age). Do you notice anything unusual about the
results?
$ cat /user/user01/7/OUT/part-r-00000
democrat

18.0

democrat

77.0

democrat

47.0

green 18.0
green 77.0

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

green 47.0
independent 18.0
independent 77.0
independent 47.0
libertarian 18.0
libertarian 77.0
libertarian 47.0

republican

18.0

republican

77.0

republican

47.0

socialist

18.0

socialist

77.0

socialist

47.0

Lesson 8: Launching Jobs

The goal of the exercise in the lessons lab is to use the tools and methods in lecture to be able to launch
multiple jobs. This lab has only one exercise with several tasks.
Exercise

Duration

8.1: Write a MapReduce Driver to launch two jobs

30 min

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Exercise 8.1: Write a MapReduce Driver to Launch Two Jobs

In this exercise, you will modify a MapReduce driver that launches two jobs. The first job calculates
minimum, maximum, and mean values for the SAT verbal and math scores. The second job calculates
the numerator and denominator for the Spearman correlation coefficient between the verbal and math
scores. The driver then calculates the correlation coefficient by dividing the numerator by the square
root of the denominator. The code for both MapReduce jobs has been provided.

Pearson's Correlation Coefficient

This statistic is used to determine the level of dependence (or correlation) between two variables. The
value of the coefficient ranges from -1 to +1, where

+1 directly proportional (high positive correlation)

0 no correlation

-1 inversely proportional (high negative correlation)

The formula to calculate this coefficient is given here:

In words, the numerator is the sum of the products of the differences between each value and its mean.
The denominator is the square root of the sum of the squares of the differences between each value
and its mean.

Programming Objective
Let X represent the SAT verbal scores and Y represent the SAT math scores. The first MapReduce job
calculates the mean values for X and Y, and the second MapReduce job calculates the numerator and
the squared value of the denominator. The driver you write must configure and launch both jobs and
then calculate the Spearman correlation coefficient.

Copy Lab Files to Cluster

In this task, you will copy the lab files to your cluster and extract them in a folder for use in this exercise.
1. Create a directory for this exercise.
$ mkdir /user/user01/8
2. Copy the lab ZIP file to your cluster.
$ scp STATISTICS_SOLUTION.zip user01@node-IP:/user/user01/8

PROPRIETARY AND CONFIDENTIAL INFORMATION
2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

3. Extract the ZIP file.

$ cd /user/user01/8
$ unzip STATISTICS_SOLUTION.zip

Implement the TODOs in the WholeJobDriver.java File

In this task, you will configure the WholeJobDriver.java file to configure and launch the two MapReduce
jobs and then calculate the correlation coefficient.
//TODO: set driver, mapper, and reducer classes for first job (university)
//TODO: set input format, output key, and output value classes for university job
//TODO: set input and output paths for university job
//TODO: launch the first job synchronously
//TODO: set driver, mapper, and reducer classes for the second job (stat)
//TODO: set input format, output key, and output value classes for stat job
//TODO: set input and output paths for stat job
//TODO: launch the second job synchronously
//TODO: return 0 or 1 based on the return codes from your 2 jobs

Run the MapReduce Jobs

In this task, you will launch the WholeJobDriver to calculate the Spearman correlation coefficient.
1. Run the wholejob_driver.sh script.
2. Your results should be as follows:
product_sumofsquares is 243128.0
var1_sumofsquares is 259871.0
var2_sumofsquares is 289679.0
spearman's coefficient is 0.886130250066755

PROPRIETARY AND CONFIDENTIAL INFORMATION
2015 MapR Technologies, Inc. All Rights Reserved.

Lab 9: Using non-Java languages for

MapReduce applications (Streaming)
The goal of the exercises in this lessons lab is to use non-Java programming to implement a MapReduce
jobs. This lab has one exercise. This lab is not required.
Exercise
9.1: Implement a MapReduce streaming application

Duration
30 min

Developing Hadoop Applications

Exercise 9.1: Implement a MapReduce Streaming Application

In this exercise, you will implement a MapReduce streaming application using the language of your
choice (Python or Perl). Guidance will be provided for building the application in the UNIX bash shell.
We return to the RECEIPTS data set to calculate the minimum, maximum, mean and the years
associated with the those values.

Copy the Lab Files to Your Cluster

In this task, you will copy and extract the lab files for this exercise.
1. Create a directory for this exercise.
$ mkdir /user/user01/9
2. Copy the lab ZIP file to this directory.
$ scp STREAMING_RECEIPTS.zip user01@node-IP:/user/user01/9
3. Extract the lab ZIP file.
$ cd /user/user01/9
$ unzip STREAMING_RECEIPTS.zip

Implement the Mapper

In this task, you will implement the mapper in the language of your choice (Python or Perl) based on the
following bash shell implementation of the same logic.
#!/bin/bash
while read record
do
year=`echo $record | awk '{print $1}'`
delta=`echo $record | awk '{print $4}'`
printf "summary\t%s_%s\n" "$year" "$delta"
done

Implement the Reducer

In this task, you will implement the reducer in the language of your choice (Python or Perl) based on the
following bash shell implementation of the same logic.
#!/bin/bash -x
count=0
sum=0
max=-2147483647
min=2147483647
minyear=""
maxyear=""
while read line
do

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

value=`echo $line | awk '{print $2}'`

if [ -n "$value" ]
then
year=`echo $value | awk -F_ '{print $1}'`
delta=`echo $value | awk -F_ '{print $2}'`
fi
if [ $delta -lt $min ]
then
min=$delta
minyear=$year
elif [ $delta -gt $max ]
then
max=$delta
maxyear=$year
fi
count=$(( count + 1 ))
sum=$(( sum + delta ))
done
mean=$(( sum / count ))
printf "min year is %s\n" "$minyear"
printf "min value is %s\n" "$min"
printf "max year is %s\n" "$maxyear"
printf "max value is %s\n" "$max"
printf "sum is %s\n" "$sum"
printf "count is %s\n" "$count"
printf "mean is %d\n" "$mean"

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.

Developing Hadoop Applications

Launch the Job

In this task, you'll modify a run script to match your environment and then run the MapReduce
streaming job.
1. Modify the receipts_driver.sh script to match the paths to your language choice (Python
or Perl), naming convention, and locations for input and output.
#!/usr/bin/env bash
USER=`whoami`
# 1) test map script
echo -e "1901 588 525 63 588 525 63" | ./receipts_mapper.sh | od -c
# 2) test reduce script
echo -e "summary\t1901_63" | ./receipts_reducer.sh | od -c
# 3) map/reduce on Hadoop
export JOBHOME=/user/$USER/9/STREAMING_RECEIPTS
export CONTRIB=/opt/mapr/hadoop/hadoop-0.20.2/contrib/streaming
export STREAMINGJAR=hadoop-*-streaming.jar
export THEJARFILE=$CONTRIB/$STREAMINGJAR
rm -rf $JOBHOME/OUT
hadoop1 jar $THEJARFILE \
-mapper 'receipts_mapper.sh' \
-file receipts_mapper.sh \
-reducer 'receipts_reducer.sh' \
-file receipts_reducer.sh \
-input $JOBHOME/DATA/receipts.txt \
-output $JOBHOME/OUT

2. Launch the MapReduce streaming job.
$ ./receipts_driver.sh

3. Examine the output.

$ cat /user/user01/9/OUT/part-r-00000
min year is 2009
min value is -1412688
max year is 2000
max value is 236241
sum is -10418784
count is 111
mean is -93862

PROPRIETARY AND CONFIDENTIAL INFORMATION

2015 MapR Technologies, Inc. All Rights Reserved.