Setup Hadoop(HDFS) on Mac

This tutorial will provide step by step instruction to setup HDFS on Mac OS.

Download Apache Hadoop
You can click here to download apache Hadoop 3.0.3 version or go to Apache site http://hadoop.apache.org/releases.html to download directly from there. Move the downloaded Hadoop binary to below path & extract it.

$HOME/hadoop/

There are 6 steps to complete in order setup Hadoop (HDFS)

Validate if java is installed
Setup environment variables in .profile file
Setup configuration files for local Hadoop
Setup password less ssh to localhost
Initialize Hadoop cluster by formatting HDFS directory
Starting Hadoop cluster

Each Step is described in detail below

Validating Java: Java version can be checked using below command. If java is not present or lower version(Java 8 is recommended) is installed then latest JDK can be download from Oracle site here and can be installed.
```
$ java -version

Output:
java version "1.8.0_141"
Java(TM) SE Runtime Environment (build 1.8.0_141-b15)
Java HotSpot(TM) 64-Bit Server VM (build 25.141-b15, mixed mode)
```

Set below variables in the .profile file in $HOME directory

## Set Java Home as env variable
export JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_141.jdk/Contents/Home


## Set HADOOP environment variables
export HADOOP_HOME=$HOME/hadoop/hadoop-3.0.3
export HADOOP_MAPRED_HOME=$HADOOP_HOME
export HADOOP_COMMON_HOME=$HADOOP_HOME
export HADOOP_HDFS_HOME=$HADOOP_HOME
export HADOOP_YARN_HOME=$HADOOP_HOME
export HADOOP_OPTS="-Djava.library.path=$HADOOP_HOME/lib/native"
export HADOOP_COMMON_LIB_NATIVE_DIR=$HADOOP_HOME/lib/native
export YARN_HOME=$HADOOP_HOME
export HADOOP_INSTALL=$HADOOP_HOME
export HADOOP_CONF_DIR=$HADOOP_HOME
export HADOOP_LIBEXEC_DIR=$HADOOP_HOME/libexec
export JAVA_LIBRARY_PATH=$HADOOP_HOME/lib/native:$JAVA_LIBRARY_PATH
export HADOOP_CONF_DIR=$HADOOP_HOME/etc/hadoop
export HADOOP_INSTALL=$HADOOP_HOME


## Set Path to the Hadoop Binary
export PATH=$PATH:$HADOOP_HOME/bin

Note: Path of Java Home can be determined by using below command in Terminal

/usr/libexec/java_home

In order to set HDFS, please make changes (as mentioned in detail below) in the following files under $HADOOP_HOME/etc/hadoop/

core-site.xml
hdfs-site.xml
mapred-site.xml
yarn-site.xml
hadoop-env.sh

core-site.xml : Please add below listed XML properties in $HADOOP_HOME/etc/hadoop/core-site.xml file

<!-- Place below properties inside configuration tag -->

<!-- Specify the port for hadoop hdfs -->
<property>
    <name>fs.defaultFS</name>
    <value>hdfs://localhost:9000</value>
</property>

<property>
    <name>dfs.replication</name>
    <value>1</value>
</property>

hdfs-site.xml : Please add below listed XML properties in $HADOOP_HOME/etc/hadoop/hdfs-site.xml file

<!-- Place below properties inside configuration tag -->

<!-- Directory Path where hadoop file system will be created(create these hadoop_storage directory and specify the path) -->
<property>
<name>dfs.name.dir</name>
<value>[$HOME Path]/hadoop/hadoop_storage</value>
<final>true</final>
</property>

<!-- Default block size is 128 Mb -->
<property>
<name>dfs.blocksize</name>
<value>134217728</value>
</property>

<!-- Replication factor is set as 1 as this is for local -->
<property>
<name>dfs.replication</name>
<value>1</value>
</property>

<!-- Specify a particular Temporary directory, otherwise it will try to use defalut temp dir of system -->
<property>
    <name>hadoop.tmp.dir</name>
    <value>[$HOME Path]/hadoop/tmp</value>
</property>

<!-- Below property is for hive setup -->
<property>
<name>hive.metastore.warehouse.dir</name>
<value>hdfs://master:8020/user/hive/warehouse</value>
</property>

yarn-site.xml : Please add below listed XML properties in $HADOOP_HOME/etc/hadoop/yarn-site.xml file

<!-- Place below properties inside configuration tag -->

<property>
    <name>yarn.nodemanager.aux-services</name>
    <value>mapreduce_shuffle</value>
</property>
<property>
    <name>yarn.nodemanager.env-whitelist</name>
    <value>JAVA_HOME, HADOOP_COMMON_HOME, HADOOP_HDFS_HOME, HADOOP_CONF_DIR, CLASSPATH_PREPEND_DISTCACHE, HADOOP_YARN_HOME, HADOOP_MAPRED_HOME
    </value>
</property>

<!-- This property specify to show system healthy until capacity is reached 98.5% -->
<property>
    <name>yarn.nodemanager.disk-health-checker.max-disk-utilization-per-disk-percentage
    </name>
    <value>98.5</value>
</property>

mapred-site.xml : Please add below listed XML properties in $HADOOP_HOME/etc/hadoop/mapred-site.xml file

<!-- Place below properties inside configuration tag -->

<property>
  <name>yarn.app.mapreduce.am.env</name>
  <value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
<property>
  <name>mapreduce.map.env</name>
  <value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
<property>
  <name>mapreduce.reduce.env</name>
  <value>HADOOP_MAPRED_HOME=$HADOOP_HOME</value>
</property>
<property>
  <name>mapred.job.tracker</name>
  <value>localhost:8021</value>
</property>

hadoop-env.sh : Please add below environment variables in $HADOOP_HOME/etc/hadoop/hadoop-env.sh file

# export JAVA_HOME [Same as in .profile file]
export JAVA_HOME=/Library/Java/JavaVirtualMachines/jdk1.8.0_141.jdk/Contents/Home

# Location of Hadoop.
export HADOOP_HOME=[Download hadoop Binary path]

Hadoop namenode & secondary namenode requires password less ssh to localhost in order to start. 2 things need to be done to setup password less ssh.
- Enable Remote Login in System Preference --> Sharing, get your username added in allowed user list if you are not administrator of the system.
- Generate & setup key
```
ssh-keygen -t rsa -P '' -f ~/.ssh/id_rsa
cat ~/.ssh/id_rsa.pub >> ~/.ssh/authorized_keys
```
Initialize Hadoop cluster by formatting HDFS directory[Run below commands in terminal]
```
source ~/.profile
$HADOOP_HOME/bin/hdfs namenode -format
```
Starting Hadoop cluster
- Starting both hdfs & yarn servers in single command
```
$HADOOP_HOME/sbin/start-all.sh
```
- Other Commands to start hdfs & yarn servers one by one.
```
$HADOOP_HOME/sbin/start-dfs.sh
$HADOOP_HOME/sbin/start-yarn.sh
```

Checking if all the namenode, datanode & resource manager started or not (using jps command)

$ jps

Output
2448 SecondaryNameNode
2646 ResourceManager
2311 DataNode
2746 NodeManager
2815 Jps
2207 NameNode

Health of the hadoop cluster & yarn processing can be checked on Web UI

Hadoop Health: http://localhost:9870

Yarn: http://localhost:8088/cluster

Stopping Hadoop cluster

Stopping both hdfs & yarn servers in single command
```
$HADOOP_HOME/sbin/stop-all.sh
```

Other Commands to start hdfs & yarn servers one by one.

$HADOOP_HOME/sbin/stop-dfs.sh
$HADOOP_HOME/sbin/stop-yarn.sh

Running Basic HDFS Command

Hadoop Version
```
$ hadoop version

Output: Hadoop 3.0.3
```
List all directories
```
hadoop fs -ls /
```
Creating user home directory
```
hadoop fs -mkdir -p /user/[username]
```

Copy file from local to HDFS

hadoop fs -copyFromLocal $HADOOP_HOME/etc/hadoop/core-site.xml .

hadoop fs -put $HADOOP_HOME/etc/hadoop/core-site.xml  /user/[username]/output

Checking data in the file on HDFS

hadoop fs -cat hdfs://localhost:9000/user/[username]/output/part-r-00000

How to Install Hadoop on Mac OS

Setup Hadoop(HDFS) on Mac