A script that I used to setup hadoop.

hadoop-setup.sh

#!/bin/bash

# Hadoop http download location
hadoop_http="https://archive.apache.org/dist/hadoop/core/hadoop-2.7.2/hadoop-2.7.2.tar.gz"

# Location to install hadoop
HADOOP_INSTALL=/opt/hadoop/

# User to run hadoop under
HADOOP_USER=hadoop

# hostanme or IP of master hdfs
HADOOP_HDFS_MASTER=nyancat

# hostname of mapred host
HADOOP_MAPRED_MASTER=nyancat

# location of tools
WGET=/usr/bin/wget
PRINTF=/usr/bin/printf
TAR=/usr/bin/tar
SED=/usr/bin/sed

function prereq {
	if [ -z "$(which javac)" ]; then
		$PRINTF "Install Java first!\n"
		exit
	fi

	if [ -z "ps aux | grep -v grep | grep sshd" ]; then
		$PRINTF "Install and run SSH server!\n"
		exit
	fi
}

function download_extract_hadoop {
	$PRINTF "download_extract_hadoop start\n\n"

	cd /tmp/
	$WGET $hadoop_http

	if [ $? -ne 0 ]; then
		$PRINTF "Failed to download hadoop from $hadoop_http"
		exit 1
	fi

	# the filename of downloaded archive
	HADOOP_TAR=$(basename $hadoop_http)

	# filename of extracted directory
	HADOOP_TARLESS=$($PRINTF $HADOOP_TAR | $SED 's/.tar.gz//')

	rm -rf $HADOOP_INSTALL
	mkdir $HADOOP_INSTALL

	$TAR -xzvf $HADOOP_TAR
	
	if [ -z "$HADOOP_TARLESS" ]; then
		$PRINTF "Variable HADOOP_TARLESS is empty!!\n"
		exit -1
	fi

	mv $HADOOP_TARLESS/* $HADOOP_INSTALL

	HADOOP_ENV_FILE="$HADOOP_INSTALL/etc/hadoop/hadoop-env.sh"
	backup_file $HADOOP_ENV_FILE
	$SED -i "s#\${JAVA_HOME}#$(dirname $(dirname $(readlink -f $(which javac))))#" $HADOOP_ENV_FILE

	$PRINTF "download_extract_hadoop done...\n\n"
}

function create_user {
	$PRINTF "create_user start\n\n"

	useradd -m $HADOOP_USER

	if [ $? -ne 0 ]; then
		$PRINTF "Failed to create user $HADOOP_USER"
		exit 1
	fi

	touch /home/$HADOOP_USER/.bashrc
	$PRINTF "export PATH=$PATH:$HADOOP_INSTALL/bin/:$HADOOP_INSTALL/sbin/:\n" >> /home/$HADOOP_USER/.bashrc
	$PRINTF "export JAVA_HOME=$(dirname $(dirname $(readlink -f $(which javac))))\n" >> /home/$HADOOP_USER/.bashrc
	export PATH=$PATH:$HADOOP_INSTALL/bin/:$HADOOP_INSTALL/sbin/:
	export JAVA_HOME=$(dirname $(dirname $(readlink -f $(which javac))))

	chown -R $HADOOP_USER /home/$HADOOP_USER/
	chown -R $HADOOP_USER $HADOOP_INSTALL
	$PRINTF "create_user done...\n\n"
}

function run_standalone {
	$PRINTF "standalone start\n\n"

	cd /home/$HADOOP_USER
	rm -rf input
	mkdir input
	cp $HADOOP_INSTALL/etc/hadoop/*.xml input
	hadoop jar $HADOOP_INSTALL/share/hadoop/mapreduce/hadoop-mapreduce-examples-2.7.2.jar grep input output 'dfs[a-z.]+'
	cat output/*
	read -rsp $'Check  output above.\n\nPress any key to continue...\n' -n1 key

	$PRINTF "standalone done...\n\n"
}

function configure_psu {
	$PRINTF "configure_psu start\n\n"

	mkdir -p /home/$HADOOP_USER/fs/{namenode,datanode,namesecondary}
	chown -R $HADOOP_USER /home/$HADOOP_USER/

	HADOOP_CORE_SITE_FILE=$HADOOP_INSTALL/etc/hadoop/core-site.xml
	backup_file $HADOOP_CORE_SITE_FILE

	HADOOP_CORE_SITE="<property>\n\
        	<name>fs.defaultFS</name>\n\
        	<value>hdfs://$HADOOP_HDFS_MASTER:8020</value>\n\
    	</property>\n\
	<property>\n\
		<name>io.file.buffer.size</name>\n\
		<value>131072</value>\n\
	</property>"
	$SED -i "/<configuration>/a\ $HADOOP_CORE_SITE" $HADOOP_CORE_SITE_FILE

	HADOOP_YARN_SITE_FILE=$HADOOP_INSTALL/etc/hadoop/yarn-site.xml
	backup_file $HADOOP_YARN_SITE_FILE

	HADOOP_YARN_SITE="<property>\n\
		<name>yarn.resourcemanager.hostname</name>\n\
		<value>$HADOOP_MAPRED_MASTER</value>\n\
	</property>\n\
	<property>\n\
		<name>yarn.nodemanager.aux-services</name>\n\
      		<value>mapreduce_shuffle</value>\n\
	</property>"

	$SED -i "/<configuration>/a\ $HADOOP_YARN_SITE" $HADOOP_YARN_SITE_FILE

	HADOOP_HDFS_SITE_FILE=$HADOOP_INSTALL/etc/hadoop/hdfs-site.xml
	backup_file $HADOOP_HDFS_SITE_FILE

	HADOOP_HDFS_SITE="<property>\n\
        <name>dfs.replication</name>\n\
        <value>3</value>\n\
    </property>\n\
	<property>\n\
		<name>dfs.namenode.name.dir</name>\n\
		<value>file:/home/$HADOOP_USER/fs/namenode</value>\n\
	</property>\n\
	<property>\n\
		<name>dfs.datanode.data.dir</name>\n\
		<value>file:/home/$HADOOP_USER/fs/datanode</value>\n\
	</property>\n\
	<property>\n\
		<name>dfs.namenode.checkpoint.dir</name>\n\
		<value>file:/home/$HADOOP_USER/fs/namesecondary</value>\n\
	</property>\n\
	<property>\n\
		<name>dfs.block.size</name>\n\
		<value>134217728</value>\n\
	</property>"
	$SED -i "/<configuration>/a\ $HADOOP_HDFS_SITE" $HADOOP_HDFS_SITE_FILE

	HADOOP_MAPRED_FILE=$HADOOP_INSTALL/etc/hadoop/mapred-site.xml
	mv $HADOOP_MAPRED_FILE.template $HADOOP_MAPRED_FILE
	backup_file $HADOOP_MAPRED_FILE
	
	HADOOP_MAPRED="<property>\n\
		<name>mapred.job.tracker</name>\n\
		<value>$HADOOP_MAPRED_MASTER:54311</value>\n\
	</property>\n\
	<property>\n\
		<name>mapreduce.framework.name</name>\n\
		<value>yarn</value>\n\
	</property>\n\
	<property>\n\
		<name>mapreduce.jobhistory.address</name>\n\
		<value>$HADOOP_MAPRED_MASTER:10020</value>\n\
	</property>\n\
	<property>\n\
		<name>mapreduce.jobhistory.webapp.address</name>\n\
		<value>$HADOOP_MAPRED_MASTER:19888</value>\n\
	</property>"
	$SED -i "/<configuration>/a\ $HADOOP_MAPRED" $HADOOP_MAPRED_FILE
	
	$PRINTF "$HADOOP_MAPRED_MASTER" > $HADOOP_INSTALL/etc/hadoop/masters
	chown -R $HADOOP_USER /home/$HADOOP_USER/

	$PRINTF "configure_psu done...\n\n"
}

function backup_file {
	if [ -z "$1" ]; then
		$PRINTF "backup_file arguement is empty!\n\n"
		return
	fi	

	cp $1 $1.original
}

function todo_pubkey {
	$PRINTF "Make this account, and $HADOOP_USER, authenticate with public keys.\n\n"
	read -rsp $'Press any key to continue...\n' -n1 key
}

function to_do {
	$PRINTF "TO DO:\n\n"
	$PRINTF "Generate ssh keys for pub key auth on master to slave.\n\n"

	chown -R $HADOOP_USER $HADOOP_INSTALL
}

prereq
download_extract_hadoop
create_user
run_standalone
configure_psu
todo_pubkey
hdfs namenode -format	

to_do