Big data processing using Hadoop with Cloudera Quickstart

thanachart@imcinstitute.com1
Big Data Processing
Using Cloudera Quickstart
with a Docker Container
July 2016
Dr.Thanachart Numnonda
IMC Institute
thanachart@imcinstitute.com
Modifiy from Original Version by Danairat T.
Certified Java Programmer, TOGAF – Silver
danairat@gmail.com

Thanachart Numnonda, thanachart@imcinstitute.com July 2016Hadoop Workshop using Cloudera on Amazon EC2
Outline
●
Launch AWS EC2 Instance
●
Install Docker on Ubuntu
●
Pull Cloudera QuickStart to the docker
●
HDFS
●
HBase
●
MapReduce
●
Hive
●
Pig
●
Impala
●
Sqoop
Hive.apache.org

Cloudera VM
This lab will use a EC2 virtual server on AWS to install
Cloudera, However, you can also use Cloudera QuickStart VM
which can be downloaded from:
http://www.cloudera.com/content/www/en-us/downloads.html

Hands-On: Launch a virtual server
on EC2 Amazon Web Services
(Note: You can skip this session if you use your own
computer or another cloud service)

Virtual Server
This lab will use a EC2 virtual server to install a
Cloudera Cluster using the following features:
Ubuntu Server 14.04 LTS
Four m3.xLarge 4vCPU, 15 GB memory,80 GB
SSD
Security group: default
Keypair: imchadoop

Select a EC2 service and click on Lunch Instance

Select an Amazon Machine Image (AMI) and
Ubuntu Server 14.04 LTS (PV)

Choose m3.xlarge Type virtual server

Add Storage: 80 GB

Name the instance

Select Create an existing security group > Default

Click Launch and choose imchadoop as a key pair

Review an instance and rename one instance as a
master / click Connect for an instruction to connect
to the instance

Connect to an instance from Mac/Linux

Can also view details of the instance such as Public
IP and Private IP

Connect to an instance from Windows using Putty

Connect to the instance

Hands-On: Installing Cloudera
Quickstart on Docker Container

Installation Steps
●
Update OS
●
Install Docker
●
Pull Cloudera Quickstart
●
Run Cloudera Quickstart
●
Run Cloudera Manager
Hive.apache.org

Update OS (Ubuntu)
●
Command: sudo apt-get update

Docker Installation
●
Command: sudo apt-get install docker.io

Pull Cloudera Quickstart
●
Command: sudo docker pull cloudera/quickstart:latest

Show docker images
●
Command: sudo docker images

Run Cloudera quickstart
●
Command: sudo docker run
--hostname=quickstart.cloudera --privileged=true -t -i
[OPTIONS] [IMAGE] /usr/bin/docker-quickstart
Example: sudo docker run
--hostname=quickstart.cloudera --privileged=true -t -i -p
8888:8888 cloudera/quickstart /usr/bin/docker-quickstart

Finding the EC2 instance's DNS

Login to Hue
http://ec2-54-173-154-79.compute-1.amazonaws.com:8888

Hadoop File System (HDFS)
IMC Institute

HDFS
●
Default storage for the Hadoop cluster
●
Data is distributed and replicated over multiple machines
●
Designed to handle very large files with straming data
access patterns.
●
NameNode/DataNode
●
Master/slave architecture (1 master 'n' slaves)
●
Designed for large files (64 MB default, but configurable)
across all the nodes
Hive.apache.org

HDFS Architecture
Source Hadoop: Shashwat Shriparv

Data Replication in HDFS
Source Hadoop: Shashwat Shriparv

How does HDFS work?
Source Introduction to Apache Hadoop-Pig: PrashantKommireddi

Hands-On: Importing/Exporting
Data to HDFS

Review file in Hadoop HDFS using
File Browse

Create a new directory name as: input & output

Upload a local file to HDFS

Hands-On: Connect to a master node
via SSH

SSH Login to a master node

Hadoop syntax for HDFS

Install wget
●
Command: yum install wget

Download an example text file
Make your own durectory at a master node to avoid mixing with others
$mkdir guest1
$cd guest1
$wget https://s3.amazonaws.com/imcbucket/input/pg2600.txt

Upload Data to Hadoop
$hadoop fs -ls /user/cloudera/input
$hadoop fs -rm /user/cloudera/input/*
$hadoop fs -put pg2600.txt /user/cloudera/input/
$hadoop fs -ls /user/cloudera/input

Lecture
Understanding HBase

Introduction
An open source, non-relational, distributed database
HBase is an open source, non-relational, distributed database
modeled after Google's BigTable and is written in Java. It is
developed as part of Apache Software Foundation's Apache
Hadoop project and runs on top of HDFS (, providing
BigTable-like capabilities for Hadoop. That is, it provides a
fault-tolerant way of storing large quantities of sparse data.

HBase Features
●
Hadoop database modelled after Google's Bigtable
●
Column oriented data store, known as Hadoop Database
●
Support random realtime CRUD operations (unlike
HDFS)
●
No SQL Database
●
Opensource, written in Java
●
Run on a cluster of commodity hardware
Hive.apache.org

When to use HBase?
●
When you need high volume data to be stored
●
Un-structured data
●
Sparse data
●
Column-oriented data
●
Versioned data (same data template, captured at various
time, time-elapse data)
●
When you need high scalability
Hive.apache.org

Which one to use?
●
HDFS
●
Only append dataset (no random write)
●
Read the whole dataset (no random read)
●
HBase
●
Need random write and/or read
●
Has thousands of operation per second on TB+ of data
●
RDBMS
●
Data fits on one big node
●
Need full transaction support
●
Need real-time query capabilities
Hive.apache.org

HBase Components
Hive.apache.org
●
Region
●
Row of table are stores
●
Region Server
●
Hosts the tables
●
Master
●
Coordinating the Region
Servers
●
ZooKeeper
●
HDFS
●
API
●
The Java Client API

HBase Shell Commands
Hive.apache.org

Hands-On: Running HBase

Hbase shell
$hbase shell
hbase(main):001:0> create 'employee', 'personal data',
'professional data'
hbase(main):002:0> list

Create Data

Running HBase Browser

Viewing Employee Table

Create a table in HBase

Insert a new row in a table

Add field into a new row

Lecture: Understanding Map Reduce
Processing
Client
Name Node Job Tracker
Data Node
Task Tracker
Data Node
Task Tracker
Data Node
Task Tracker
Map Reduce

Hadoop Ecosystem

Source: The evolution and future of Hadoop storage: Cloudera

Before MapReduce…
●
Large scale data processing was difficult!
– Managing hundreds or thousands of processors
– Managing parallelization and distribution
– I/O Scheduling
– Status and monitoring
– Fault/crash tolerance
●
MapReduce provides all of these, easily!
Source: http://labs.google.com/papers/mapreduce-osdi04-slides/index-auto-0002.html

MapReduce Framework
Source: www.bigdatauniversity.com

How Map and Reduce Work Together
●
Map returns information
●
Reduces accepts information
●
Reduce applies a user defined function to reduce the
amount of data

Map Abstraction
●
Inputs a key/value pair
– Key is a reference to the input value
– Value is the data set on which to operate
●
Evaluation
– Function defined by user
– Applies to every value in value input
●
Might need to parse input
●
Produces a new list of key/value pairs
– Can be different type from input pair

Reduce Abstraction
●
Starts with intermediate Key / Value pairs
●
Ends with finalized Key / Value pairs
●
Starting pairs are sorted by key
●
Iterator supplies the values for a given key to the
Reduce function.

Reduce Abstraction
●
Typically a function that:
– Starts with a large number of key/value pairs
●
One key/value for each word in all files being greped
(including multiple entries for the same word)
– Ends with very few key/value pairs
●
One key/value for each unique word across all the files with
the number of instances summed into this entry
●
Broken up so a given worker works with input of the
same key.

Why is this approach better?
●
Creates an abstraction for dealing with complex
overhead
– The computations are simple, the overhead is messy
●
Removing the overhead makes programs much
smaller and thus easier to use
– Less testing is required as well. The MapReduce
libraries can be assumed to work properly, so only
user code needs to be tested
●
Division of labor also handled by the
MapReduce libraries, so programmers only
need to focus on the actual computation

Hands-On: Writing you own Map
Reduce Program

Example MapReduce: WordCount

Running Map Reduce Program
$cd /root/guest1
$wget https://dl.dropboxusercontent.com/u/12655380/wordcount.jar
$hadoop jar wordcount.jar org.myorg.WordCount
/user/cloudera/input/* /user/cloudera/output/wordcount

Reviewing MapReduce Job in Hue

Reviewing MapReduce Output Result

Lecture
Understanding Oozie

Introduction
Workslow scheduler for Hadoop
Oozie is a workflow scheduler system to manage Apache
Hadoop jobs. Oozie is integrated with the rest of the Hadoop
stack supporting several types of Hadoop jobs out of the box
(such as Java map-reduce, Streaming map-reduce, Pig, Hive,
Sqoop and Distcp) as well as system specific jobs (such as
Java programs and shell scripts).

What is Oozie?
●
Work flow scheduler for Hadoop
●
Manages Hadoop Jobs
●
Integrated with many Hadoop apps i.e. Pig, Hive
●
Scaleable
●
Schedule jobs
●
A work flow is a collection of actions.
●
A work flow is
– Arranged as a DAG ( direct acyclic graph )
– Graph stored as hPDL ( XML process definition )

Oozie Architecture
Source: info@semtech-solutions.co.nz

Oozie Server
Source: Oozie – Now and Beyond, Yahoo, 2013

Layer of Abstraction in Oozie

Workflow Example: Data Analytics
●
Logs => fact table(s)
●
Database backup => Dimension tables
●
Complete rollups/cubes
●
Load data into a low-latency storage (e.g. Hbae, HDFS)
●
Dashboard & BI tools
Source: Workflow Engines for Hadoop, Joe Crobak, 2013

●
What happens if there is a failure?
– Rebuild the failed day
– .. and any downstream datasets
●
With Hadoop Workflow
– Possible OK to skip a day
– Workflow tends to be self-contained, so you do not need to run
downstream.
– Sanity check your data before pushing to production.

Oozie Workflow

Oozie Use Cases
●
Time Triggers
– Execute your workflow every 15 minutes
●
Time and Data Triggers
– Materialize your workflow every hour, but only run them when
the input data is ready (that is loaded to the grid every hour)
●
Rolling Window
– Access 15 minute datasets and roll them up into hourly
datasets

Hands-On: Running Map Reduce
using Oozie workflow

Using Hue: select WorkFlows >> Editors >>
Workflows

Create a new workflow
●
Click Create button; the following screen will be displayed
●
Name the workflow as WordCountWorkflow

Select a Java job for the workflow
●
From the Oozie editor, drag Java Program and drop between start and
end

Edit the Java Job
●
Assign the following value
–
– Jar name: wordcount.jar (select … choose upload from local
machine)
– Main Class: org.myorg.WordCount
– Arguments: /user/cloudera/input/*
– /user/cloudera/output/wordcount

Submit the workflow
●
Click Done, follow by Save
●
Then click submit

Introduction
A Petabyte Scale Data Warehouse Using Hadoop
Hive is developed by Facebook, designed to enable easy data
summarization, ad-hoc querying and analysis of large
volumes of data. It provides a simple query language called
Hive QL, which is based on SQL

Danairat T., danairat@gmail.com: Thanachart N., thanachart@imcinstitute.com April 2015Big Data Hadoop Workshop
What Hive is NOT
Hive is not designed for online transaction processing and
does not offer real-time queries and row level updates. It is
best used for batch jobs over large sets of immutable data
(like web logs, etc.).

Sample HiveQL
The Query compiler uses the information stored in the metastore to
convert SQL queries into a sequence of map/reduce jobs, e.g. the
following query
SELECT * FROM t where t.c = 'xyz'
SELECT t1.c2 FROM t1 JOIN t2 ON (t1.c1 = t2.c1)
SELECT t1.c1, count(1) from t1 group by t1.c1
Hive.apache.or
g

System Architecture and Components
Metastore: To store the meta data.
Query compiler and execution engine: To convert SQL queries to a
sequence of map/reduce jobs that are then executed on Hadoop.
SerDe and ObjectInspectors: Programmable interfaces and
implementations of common data formats and types.
A SerDe is a combination of a Serializer and a Deserializer (hence, Ser-De). The Deserializer interface takes a string or
binary representation of a record, and translates it into a Java object that Hive can manipulate. The Serializer, however,
will take a Java object that Hive has been working with, and turn it into something that Hive can write to HDFS or
another supported system.
UDF and UDAF: Programmable interfaces and implementations for
user defined functions (scalar and aggregate functions).
Clients: Command line client similar to Mysql command line.
hive.apache.or
g

Architecture Overview
HDFS
Hive CLI
Querie
s
Browsin
g
Map Reduce
MetaStore
Thrift API
SerDe
Thrift Jute JSON..
Execution
Hive QL
Parser
Planner
Mgmt.
WebUI
HDFS
DDL
Hive
Hive.apache.org

Hive Metastore
Hive Metastore is a repository to keep all Hive
metadata; Tables and Partitions definition.
By default, Hive will store its metadata in
Derby DB

Hive Built in Functions
Return Type Function Name (Signature) Description
BIGINT round(double a) returns the rounded BIGINT value of the double
BIGINT floor(double a) returns the maximum BIGINT value that is equal or less than the double
BIGINT ceil(double a) returns the minimum BIGINT value that is equal or greater than the double
double rand(), rand(int seed)
returns a random number (that changes from row to row). Specifiying the seed will make sure the generated
random number sequence is deterministic.
string concat(string A, string B,...)
returns the string resulting from concatenating B after A. For example, concat('foo', 'bar') results in 'foobar'. This
function accepts arbitrary number of arguments and return the concatenation of all of them.
string substr(string A, int start)
returns the substring of A starting from start position till the end of string A. For example, substr('foobar', 4) results
in 'bar'
string substr(string A, int start, int length) returns the substring of A starting from start position with the given length e.g. substr('foobar', 4, 2) results in 'ba'
string upper(string A)
returns the string resulting from converting all characters of A to upper case e.g. upper('fOoBaR') results in
'FOOBAR'
string ucase(string A) Same as upper
string lower(string A) returns the string resulting from converting all characters of B to lower case e.g. lower('fOoBaR') results in 'foobar'
string lcase(string A) Same as lower
string trim(string A) returns the string resulting from trimming spaces from both ends of A e.g. trim(' foobar ') results in 'foobar'
string ltrim(string A)
returns the string resulting from trimming spaces from the beginning(left hand side) of A. For example, ltrim(' foobar
') results in 'foobar '
string rtrim(string A)
returns the string resulting from trimming spaces from the end(right hand side) of A. For example, rtrim(' foobar ')
results in ' foobar'
string
regexp_replace(string A, string B,
string C)
returns the string resulting from replacing all substrings in B that match the Java regular expression syntax(See
Java regular expressions syntax) with C. For example, regexp_replace('foobar', 'oo|ar', ) returns 'fb'
string from_unixtime(int unixtime)
convert the number of seconds from unix epoch (1970-01-01 00:00:00 UTC) to a string representing the timestamp
of that moment in the current system time zone in the format of "1970-01-01 00:00:00"
string to_date(string timestamp) Return the date part of a timestamp string: to_date("1970-01-01 00:00:00") = "1970-01-01"
int year(string date)
Return the year part of a date or a timestamp string: year("1970-01-01 00:00:00") = 1970, year("1970-01-01") =
1970
int month(string date)
Return the month part of a date or a timestamp string: month("1970-11-01 00:00:00") = 11, month("1970-11-01") =
11
int day(string date) Return the day part of a date or a timestamp string: day("1970-11-01 00:00:00") = 1, day("1970-11-01") = 1
string
get_json_object(string json_string,
string path)
Extract json object from a json string based on json path specified, and return json string of the extracted json
object. It will return null if the input json string is invalid
hive.apache.org

Hive Aggregate Functions
Return Type
Aggregation Function Name
(Signature)
Description
BIGINT
count(*), count(expr),
count(DISTINCT expr[,
expr_.])
count(*) - Returns the total number of retrieved rows,
including rows containing NULL values; count(expr) - Returns
the number of rows for which the supplied expression is non-
NULL; count(DISTINCT expr[, expr]) - Returns the number of
rows for which the supplied expression(s) are unique and
non-NULL.
DOUBLE
sum(col), sum(DISTINCT
col)
returns the sum of the elements in the group or the sum of
the distinct values of the column in the group
DOUBLE avg(col), avg(DISTINCT col)
returns the average of the elements in the group or the
average of the distinct values of the column in the group
DOUBLE min(col) returns the minimum value of the column in the group
DOUBLE max(col) returns the maximum value of the column in the group
hive.apache.org

Running Hive
Hive Shell
Interactive
hive
Script
hive -f myscript
Inline
hive -e 'SELECT * FROM mytable'
Hive.apache.or
g

Hive Commands
ortonworks.com

: Thanachart N., thanachart@imcinstitute.com April 2015Big Data Hadoop Workshop
Hive Tables
●
Managed- CREATE TABLE
●
LOAD- File moved into Hive's data warehouse directory
●
DROP- Both data and metadata are deleted.
●
External- CREATE EXTERNAL TABLE
●
LOAD- No file moved
●
DROP- Only metadata deleted
●
Use when sharing data between Hive and Hadoop applications
or you want to use multiple schema on the same data

Hive External Table
Dropping External Table using Hive:-
Hive will delete metadata from metastore
Hive will NOT delete the HDFS file
You need to manually delete the HDFS file

Java JDBC for Hive
import java.sql.SQLException;
import java.sql.Connection;
import java.sql.ResultSet;
import java.sql.Statement;
import java.sql.DriverManager;

public class HiveJdbcClient {
  private static String driverName = "org.apache.hadoop.hive.jdbc.HiveDriver";

  public static void main(String[] args) throws SQLException {
    try {
      Class.forName(driverName);
    } catch (ClassNotFoundException e) {
      // TODO Auto-generated catch block
      e.printStackTrace();
      System.exit(1);
    }
    Connection con = DriverManager.getConnection("jdbc:hive://localhost:10000/default", "", "");
    Statement stmt = con.createStatement();
    String tableName = "testHiveDriverTable";
    stmt.executeQuery("drop table " + tableName);
    ResultSet res = stmt.executeQuery("create table " + tableName + " (key int, value string)");
    // show tables
    String sql = "show tables '" + tableName + "'";
    System.out.println("Running: " + sql);
    res = stmt.executeQuery(sql);
    if (res.next()) {
      System.out.println(res.getString(1));
    }
    // describe table
    sql = "describe " + tableName;
    System.out.println("Running: " + sql);
    res = stmt.executeQuery(sql);
    while (res.next()) {
      System.out.println(res.getString(1) + "t" + res.getString(2));
    }

HiveQL and MySQL Comparison
ortonworks.com

HiveQL and MySQL Query Comparison
ortonworks.com

Hands-On: Loading Data using Hive

hive> quit;
Quit from Hive
Start Hive

See also: https://cwiki.apache.org/Hive/languagemanual-ddl.html
Create Hive Table

Reviewing Hive Table in HDFS

Alter and Drop Hive Table
Hive > alter table test_tbl add columns (remarks STRING);
hive > describe test_tbl;
OK
id int
country string
remarks string
Time taken: 0.077 seconds
hive > drop table test_tbl;
OK
Time taken: 0.9 seconds
See also: https://cwiki.apache.org/Hive/adminmanual-metastoreadmin.html

Preparing Large Dataset
http://grouplens.org/datasets/movielens/

MovieLen Dataset
1)Type command > wget
http://files.grouplens.org/datasets/movielens/ml-100k.zip
2)Type command > yum install unzip
3)Type command > unzip ml-100k.zip
4)Type command > more ml-100k/u.user

Moving dataset to HDFS
1)Type command > cd ml-100k
2)Type command > hadoop fs -mkdir /user/cloudera/movielens
3)Type command > hadoop fs -put u.user /user/cloudera/movielens
4)Type command > hadoop fs -ls /user/cloudera/movielens

CREATE & SELECT Table

Bay Area Bike Share (BABS)
http://www.bayareabikeshare.com/open-data

Preparing a bike data
$wget https://s3.amazonaws.com/babs-open-data/
babs_open_data_year_1.zip
$unzip babs_open_data_year_1.zip
$cd 201402_babs_open_data/
$hadoop fs -put 201402_trip_data.csv
/user/cloudera
$ hadoop fs -ls /user/cloudera

Importing CSV Data with the Metastore App
The BABS data set contains 4 CSVs that contain data for
stations, trips, rebalancing (availability), and weather. We will
import trips dataset using Metastore Tables

Select: Create a new table from a file

Name a table and select a file

Choose Delimiter

Define Column Types

Create Table : Done

Starting Hive Editor

Find the top 10 most popular start stations
based on the trip data
SELECT startterminal, startstation, COUNT(1) AS count FROM trip
GROUP BY startterminal, startstation ORDER BY count DESC LIMIT 10

Introduction
A high-level platform for creating MapReduce programs Using Hadoop
Pig is a platform for analyzing large data sets that consists of
a high-level language for expressing data analysis programs,
coupled with infrastructure for evaluating these programs.
The salient property of Pig programs is that their structure is
amenable to substantial parallelization, which in turns enables
them to handle very large data sets.

Pig Components
●
Two Compnents
●
Language (Pig Latin)
●
Compiler
●
Two Execution Environments
●
Local
pig -x local
●
Distributed
pig -x mapreduce
Hive.apache.org

Running Pig
●
Script
pig myscript
●
Command line (Grunt)
pig
●
Embedded
Writing a java program
Hive.apache.org

Pig Latin
Hive.apache.org

Pig Execution Stages
Hive.apache.orgSource Introduction to Apache Hadoop-Pig: PrashantKommireddi

Why Pig?
●
Makes writing Hadoop jobs easier
●
5% of the code, 5% of the time
●
You don't need to be a programmer to write Pig scripts
●
Provide major functionality required for
DatawareHouse and Analytics
●
Load, Filter, Join, Group By, Order, Transform
●
User can write custom UDFs (User Defined Function)
Hive.apache.orgSource Introduction to Apache Hadoop-Pig: PrashantKommireddi

Pig v.s. Hive
Hive.apache.org

Hands-On: Running a Pig script

Starting Pig Command Line
$ pig -x mapreduce
2013-08-01 10:29:00,027 [main] INFO org.apache.pig.Main - Apache Pig
version 0.11.1 (r1459641) compiled Mar 22 2013, 02:13:53
2013-08-01 10:29:00,027 [main] INFO org.apache.pig.Main - Logging error
messages to: /home/hdadmin/pig_1375327740024.log
2013-08-01 10:29:00,066 [main] INFO org.apache.pig.impl.util.Utils -
Default bootup file /home/hdadmin/.pigbootup not found
2013-08-01 10:29:00,212 [main] INFO
org.apache.pig.backend.hadoop.executionengine.HExecutionEngine - Connecting
to hadoop file system at: file:///
grunt>

Writing a Pig Script for wordcount
A = load '/user/cloudera/input/*';
B = foreach A generate flatten(TOKENIZE((chararray)$0)) as
word;
C = group B by word;
D = foreach C generate COUNT(B), group;
store D into '/user/cloudera/output/wordcountPig';

Impala
IMC Institute

Introduction
open source massively parallel processing (MPP) SQL query engine
Cloudera Impala is a query engine that runs on Apache
Hadoop. Impala brings scalable parallel database technology
to Hadoop, enabling users to issue low-latency SQL queries to
data stored in HDFS and Apache HBase without requiring data
movement or transformation.

What is Impala?
General--- purpose SQL engine
Real--time queries in Apache Hadoop
Opensource under Apache License
Runs directly within Hadoop
High performance
– C++ instead of Java
– Runtime code generator
– Roughly 4-100 x Hive

Impala Overview
Impala daemon run on HDFS nodes
Statestore (for cluster metadata) v.s. Metastore (for
database metastore)
Queries run on “revelants” nodes
Support common HDFS file formats
Submit quries via Hue/Beeswax
No fault tolerant

Impala Architecture

Start Impala Query Editor

Update the list of tables/metadata by excute
the command invalidate metadata

Restart Impala Query Editor and refresh the
table list

Find the top 10 most popular start stations
based on the trip data: Using Impala
SELECT startterminal, startstation, COUNT(1) AS count FROM trip
GROUP BY startterminal, startstation ORDER BY count DESC LIMIT 10

Find the total number of trips and average duration
(in minutes) of those trips, grouped by hour
SELECT
hour,
COUNT(1) AS trips,
ROUND(AVG(duration) / 60) AS avg_duration
FROM (
SELECT
CAST(SPLIT(SPLIT(t.startdate, ' ')[1], ':')[0] AS INT) AS
hour,
t.duration AS duration
FROM `bikeshare`.`trips` t
WHERE
t.startterminal = 70
AND
t.duration IS NOT NULL
) r
GROUP BY hour
ORDER BY hour ASC;

Apache Sqoop
IMC Institute

Introduction
Sqoop (“SQL-to-Hadoop”) is a straightforward command-line
tool with the following capabilities:
●
Imports individual tables or entire databases to files in HDFS
●
Generates Java classes to allow you to interact with your
imported data
●
Provides the ability to import from SQL databases straight
into your Hive data warehouse
See also: http://sqoop.apache.org/docs/1.4.2/SqoopUserGuide.html

Hive.apache.org

Sqoop Benefit
Leverages RDBMS metadata to get the column data
types
It is simple to script and uses SQL
It can be used to handle change data capture by
importing daily transactional data to Hadoop
It uses MapReduce for export and import that enables
parallel and efficient data movement

Sqoop Mode
Sqoop import: Data moves from RDBMS to Hadoop
Sqoop export: Data moves from Hadoop to RDBMS

Use Case #1: ETL for Data Warehouse
Source: Mastering Apache Sqoop, David Yahalom, 2016

Use Case #2: ELT

Use Case #3: Data Analysis

Use Case #4: Data Archival

Use Case #5: Data Consolidation

Use Case #6: Move reports to Hadoop

Import Commands

Architecture of the import process

Incremental import

Export Commands

Hands-On: Loading Data from RDBMS
to Hadoop

●
Command: sudo docker pull mysql
●
Command: sudo docker run --name imcMysql -e
MYSQL_ROOT_PASSWORD=imcinstitute -p 3306:3306
-d mysql
●
Command: sudo docker exec -it imcMysql bash
Running MySQL Docker
root@f1922a70e09c:/# mysql -uroot -p"imcinstitute"

Prepare a test database table
mysql> CREATE DATABASE imc_db;
mysql> USE imc_db;
mysql> CREATE TABLE country_tbl(id INT NOT NULL, country
VARCHAR(50), PRIMARY KEY (id));
mysql> INSERT INTO country_tbl VALUES(1, 'USA');
mysql> INSERT INTO country_tbl VALUES(2, 'CANADA');
mysql> INSERT INTO country_tbl VALUES(3, 'Mexico');
mysql> INSERT INTO country_tbl VALUES(4, 'Brazil');
mysql> INSERT INTO country_tbl VALUES(61, 'Japan');
mysql> INSERT INTO country_tbl VALUES(65, 'Singapore');
mysql> INSERT INTO country_tbl VALUES(66, 'Thailand');

View data in the table
mysql> SELECT * FROM country_tbl;
mysql> exit;
Then exit from the container by press Ctrl-P & Ctrl-Q

Command: sudo run --hostname=quickstart.cloudera
--privileged=true --link imcMysql:mysqldb -t -i -p
8888:8888 cloudera/quickstart /usr/bin/docker-quickstart
If both of these Dockers are up and running, you can find
out the internal IP address of each of them by running
this command. This gets the IP for imcMysql.
●
Command: sudo docker inspect imcMysql | grep
IPAddress
Restart the Cloudera docker with linking to the
MySQL Docker

Command: sudo docker run
--hostname=quickstart.cloudera --privileged=true --link
imcMysql:mysqldb -t -i -p 8888:8888 cloudera/quickstart
/usr/bin/docker-quickstart
If both of these Dockers are up and running, you can find
out the internal IP address of each of them by running
this command. This gets the IP for imcMysql.
●
Command: sudo docker inspect imcMysql | grep
IPAddress
Restart the Cloudera docker with linking to the
MySQL Docker

Check a MySQL driver for Sqoop
$ cd /var/lib/sqoop
$ ls
Note: If you do not see the driver file, you need to install
one by using the following command
$ wget https://s3.amazonaws.com/imcbucket/apps/mysql-connector-java-
5.1.23-bin.jar

Importing data from MySQL to HDFS
$sqoop import --connect jdbc:mysql://172.17.0.7/imc_db
--username root --password imcinstitute --table country_tbl
--target-dir /user/cloudera/testtable -m 1

Importing data from MySQL to Hive Table
--hive-import --hive-table country -m 1

Reviewing data from Hive Table
[root@quickstart /]# hive
hive> show tables;
hive> select * from country;

Running from Hue: Beewax

Importing data from MySQL to HBase
--hbase-table country --column-family hbase_country_cf --hbase-row-key
id --hbase-create-table -m 1
$hbase shell
hbase(main):001:0> list
Start HBase

Viewing Hbase data

Viewing data from Hbase browser

Apache Flume
IMC Institute

Introduction
Apache Flume is:
●
A distributed data transport and aggregation system for
event- or log-structured data
●
Principally designed for continuous data ingestion into
Hadoop… But more flexible than that

What is Flume?
●
Apache Flume is a continuous data ingestion system that
is...
●
open-source,
●
reliable,
●
scalable,
●
manageable,
●
Customizable,
●
and designed for
Big Data ecosystem

Flume Agent
Source: Using Flume, Hari Shreedharan, 2014
●
A source writes events to one or more channels.
●
A channel is the holding area as events are passed from
a source to a sink.
●
A sink receives events from one channel only.
●
An agent can have many channels.

Sources
●
Different Source types:
●
Require at least one channel to function
●
Specialized sources for integrating with well-
known systems.
●
Example: Spooling Files, Syslog, Netcat, JMS
●
Auto-Generating Sources: Exec, SEQ
●
IPC sources for Agent-to-Agent communication: Avro,
Thrift

Channel
●
Different Channels offer different levels of
persistence:
●
Memory Channel
●
File Channel:
●
Eventually, when the agent comes back data can
be accessed.
●
Channels are fully transactional
●
Provide weak ordering guarantees
●
Can work with any number of Sources and Sinks

Sink
●
Different types of Sinks:
●
Terminal sinks that deposit events to their final
destination. For example: HDFS, HBase, Morphline-
Solr, Elastic Search
●
Sinks support serialization to user’s preferred formats.
●
HDFS sink supports time-based and arbitrary
bucketing of data while writing to HDFS.
●
IPC sink for Agent-to-Agent communication: Avro,
Thrift
●
Require exactly one channel to function

Flume Process

Flow

Flume terminology
●
A source writes events to one or more channels.
●
A channel is the holding area as events are passed from
a source to a sink.
●
A sink receives events from one channel only.
●
An agent can have many channels.
Odiago

Flume Agent Configuration : Example
Odiago

Stream Processing Architecture
Odiago

Hands-On: Loading Twitter Data to
Hadoop HDFS

Exercise Overview
Hive.apache.org

Installing Pre-built version of flume
$ wget
http://files.cloudera.com/samples/flume-sources-1.0-SNAPSHOT.jar
$ sudo cp flume-sources-1.0-SNAPSHOT.jar
/opt/cloudera/parcels/CDH-5.5.1-1.cdh5.5.1.p0.11/lib/flume-
ng/lib/
$sudo cp /etc/flume-ng/conf/flume-env.sh.template
/etc/flume-ng/conf/flume-env.sh

Create a new Twitter App
Login to your Twitter @ twitter.com

Create a new Twitter App (cont.)
Create a new Twitter App @ apps.twitter.com

Enter all the details in the application:

Your application will be created:

Click on Keys and Access Tokens:

Your Access token got created:

Add classpath in Cloudera Manager
"Services" -> "flume1" -> "Configuration" -> -> "Advanced"
-> "Java Configuration Options for Flume Agent", add:
--classpath /opt/cloudera/parcels/CDH-5.5.1-
1.cdh5.5.1.p0.11/lib/flume-ng/lib/flume-sources-1.0-
SNAPSHOT.jar

Change the Flume Agent Name

Configuring the Flume Agent

Agent Configuration
TwitterAgent.sources = Twitter
TwitterAgent.channels = MemChannel
TwitterAgent.sinks = HDFS
TwitterAgent.sources.Twitter.type =
org.apache.flume.source.twitter.TwitterSource
TwitterAgent.sources.Twitter.channels = MemChannel
TwitterAgent.sources.Twitter.consumerKey =
MjpswndxVj27ylnpOoSBrnfLX
TwitterAgent.sources.Twitter.consumerSecret =
QYmuBO1smD5Yc3zE0ZF9ByCgeEQxnxUmhRVCisAvPFudYVjC4a
TwitterAgent.sources.Twitter.accessToken = 921172807-
EfMXJj6as2dFECDH1vDe5goyTHcxPrF1RIJozqgx
TwitterAgent.sources.Twitter.accessTokenSecret =
HbpZEVip3D5j80GP21a37HxA4y10dH9BHcgEFXUNcA9xy

Agent Configuration
TwitterAgent.sources.Twitter.keywords = hadoop, big data,
analytics, bigdata, cloudera, data science, data scientiest,
business intelligence, mapreduce, data warehouse, data
warehousing, mahout, hbase, nosql, newsql,
businessintelligence, cloudcomputing
TwitterAgent.sinks.HDFS.channel = MemChannel
TwitterAgent.sinks.HDFS.type = hdfs
TwitterAgent.sinks.HDFS.hdfs.path =
hdfs://xx.xx.xx.xx:8020/user/flume/tweets/
TwitterAgent.sinks.HDFS.hdfs.fileType = DataStream
TwitterAgent.sinks.HDFS.hdfs.writeFormat = Text
TwitterAgent.sinks.HDFS.hdfs.batchSize = 1000
TwitterAgent.sinks.HDFS.hdfs.rollSize = 0
TwitterAgent.sinks.HDFS.hdfs.rollCount = 10000
TwitterAgent.channels.MemChannel.type = memory
TwitterAgent.channels.MemChannel.capacity = 10000
TwitterAgent.channels.MemChannel.transactionCapacity = 100

Restart Flume

View an agent log file

View a result using Hue

Stop the agent

7. Analyse data using Hive
$ wget
http://files.cloudera.com/samples/hive-serdes-1.0-SNAPSHOT.jar
$ mv hive-serdes-1.0-SNAPSHOT.jar /usr/local/apache-hive-
1.1.0-bin/lib/
$ hive
hive> ADD JAR /usr/local/apache-hive-1.1.0-bin/lib/hive-
serdes-1.0-SNAPSHOT.jar;
Get a Serde Jar File for parsing JSON file
Register the Jar file.

ุAnalyse data using Hive (cont.)
Running the following hive command
http://www.thecloudavenue.com/2013/03/analyse-tweets-using-flume-hadoop-and.html

Analyse data using Hive (cont)
hive> select user.screen_name, user.followers_count c from
tweets order by c desc;
Finding user who has the most number of followers

Apache Kafka
IMC Institute

Introduction
Open-source message broker project
An open-source message broker project developed by the
Apache Software Foundation written in Scala. The project
aims to provide a unified, high-throughput, low-latency
platform for handling real-time data feeds. It is, in its essence,
a "massively scalable pub/sub message queue architected as
a distributed transaction log", making it highly valuable for
enterprise infrastructures.

What is Kafka?
An apache project initially developed at LinkedIn
Distributed publish-subscribe messaging system
Designed for processing of real time activity stream data
e.g. logs, metrics collections
Written in Scala
Does not follow JMS Standards, neither uses JMS APIs

Kafka: Features
Persistent messaging
High-throughput
Supports both queue and topic semantics
Uses Zookeeper for forming a cluster of nodes
(producer/consumer/broker)
and many more…

Why Kafka?
Built with speed and scalability in mind.
Enabled near real-time access to any data source
Empowered hadoop jobs
Allowed us to build real-time analytics
Vastly improved our site monitoring and alerting
capability
Enabled us to visualize and track our call graphs.

Messaging System Concept: Queue
Source: Real time Analytics with Apache Kafka and Spark, Rahul Jain

Messaging System Concept: Topic

Terminology
Kafka maintains feeds of messages in categories called
topics.
Processes that publish messages to a Kafka topic are
called producers.
Processes that subscribe to topics and process the feed
of published messages are called consumers.
Kafka is run as a cluster comprised of one or more
servers each of which is called a broker.

Kafka

Topics
Topic: feed name to which messages are published
Source: Apache Kafka with Spark Streaming - Real Time Analytics Redefined

Topics

Topics
A topic consists of partitions.
Partition: ordered + immutable sequence of messages
that is continually appended

Kafka Architecture

Hands-on
SparkStreaming with Kafka

Install & Start Kafka Server
# wget http://www-us.apache.org/dist/kafka/0.9.0.1/kafka_2.10-
0.9.0.1.tgz
# tar xzf kafka_2.10-0.9.0.1.tgz
# cd kafka_2.10-0.9.0.1
# bin/kafka-server-start.sh config/server.properties&

Running Kafka Producer
# bin/kafka-console-producer.sh --topic test --broker-list
localhost:9092
type some random messages followed by Ctrl-D to finish

Running Kafka Consumer
# bin/kafka-console-consumer.sh --topic test --zookeeper
localhost:2181 --from-beginning

Start Spark-shell with extra memory

Spark Streaming with Kafka
$ scala> :paste
import org.apache.spark.SparkConf
import org.apache.spark.streaming.{Seconds, StreamingContext}
import org.apache.spark.storage.StorageLevel
import StorageLevel._
import org.apache.spark._
import org.apache.spark.streaming._
import org.apache.spark.streaming.StreamingContext._
import org.apache.spark.streaming.kafka.KafkaUtils
val ssc = new StreamingContext(sc, Seconds(2))
val kafkaStream = KafkaUtils.createStream(ssc,
"localhost:2181","spark-streaming-consumer-group", Map("spark-
topic" -> 5))
kafkaStream.print()
ssc.start

Running Kafka Producer on another terminal
# docker ps
# docker exec -i -t c77e4dc1ed9b /bin/bash
[root@quickstart ~]# cd /root/kafka_2.10-0.9.0.1
[root@quickstart kafka_2.10-0.9.0.1]# bin/kafka-console-
producer.sh --broker-list localhost:9092 --topic spark-topic

Test & View the result
Result from another terminal

Thank you
www.imcinstitute.com
www.facebook.com/imcinstitute

Big data processing using Hadoop with Cloudera Quickstart

More Related Content

What's hot (20)

Viewers also liked (15)

Similar to Big data processing using Hadoop with Cloudera Quickstart (20)

More from IMC Institute (20)

Recently uploaded (20)

Big data processing using Hadoop with Cloudera Quickstart