Install ES and Kafka

#!/bin/bash

kafka_version=0.9.0.1
kafka_version_scala=2.11
kafka_version_full=${kafka_version_scala}-${kafka_version}
elasticsearch_version=2.2.0
elasticsearch_clustername=bnet-es
elasticsearch_nodename=bnet-es-node01
java_version=java-1.8.0-openjdk.x86_64

cd /tmp

# ensure some tools are here...
sudo yum install tar
sudo yum install wget

# install java
echo Installing Java...
sudo yum install --quiet --assumeyes ${java_version}

# install elasticsearch
# This results in Elasticsearch being installed in:
# binaries   : /usr/share/elasticsearch/
# config     : /etc/elasticsearch
# init script: /etc/init.d/elasticsearch
echo Downloading ElasticSearch...
wget -q https://download.elasticsearch.org/elasticsearch/release/org/elasticsearch/distribution/rpm/elasticsearch/${elasticsearch_version}/elasticsearch-${elasticsearch_version}.rpm
echo Installing ElasticSearch v${elasticsearch_version}...
sudo rpm -ivh elasticsearch-${elasticsearch_version}.rpm

# configure elasticsearch by prepending our settings
echo -e "node.name: ${elasticsearch_nodename}\ncluster.name: ${elasticsearch_clustername}\nnetwork.bind_host: 0\n" > /tmp/tmpfile.$$
# why bind_host:0? ==> http://stackoverflow.com/questions/33412549/how-to-bind-elasticsearch-2-0-on-both-loopback-and-non-loopback-interfaces
sudo cat /etc/elasticsearch/elasticsearch.yml >> /tmp/tmpfile.$$
sudo mv /tmp/tmpfile.$$ /etc/elasticsearch/elasticsearch.yml

# turn elasticsearch on
sudo systemctl daemon-reload
sudo systemctl start elasticsearch.service
echo ElasticSearch installed and enabled on port 9200.
echo
echo

# Kafka's turn
# it will be in /usr/share/kafka_${kafka_version}
echo Downloading Kafka ${kafka_version}...
cd /usr/share
sudo wget -q http://apache.mivzakim.net/kafka/${kafka_version}/kafka_${kafka_version_full}.tgz
sudo tar -zxvf kafka_${kafka_version_full}.tgz
sudo rm kafka_${kafka_version_full}.tgz

# create scripts to start ZK/Kafka
echo Creating Kafka start scripts...
sudo touch /usr/bin/start-{zookeeper,kafka}
sudo chmod 755 /usr/bin/start-{zookeeper,kafka}


cat <<EOH | sudo tee /usr/bin/start-zookeeper >/dev/null
/usr/share/kafka_${kafka_version_full}/bin/zookeeper-server-start.sh \
  /usr/share/kafka_${kafka_version_full}/config/zookeeper.properties &

EOH


cat <<EOH | sudo tee /usr/bin/start-kafka >/dev/null
start-zookeeper
sleep 3
/usr/share/kafka_${kafka_version_full}/bin/kafka-server-start.sh \
  /usr/share/kafka_${kafka_version_full}/config/server.properties &

EOH

echo Use sudo start-kafka to start Kafka. Starting the first time for you now.
sudo start-kafka

You can save a bit of time by combining your yum installs (so you don't have to repeatedly check mirrors and such):

sudo yum install --quiet --assumeyes tar wget ${java_version}

---

You can pass the url for ES's RPM directly to rpm and it'll download it for you, so you don't have to worry about being in /tmp or leaving a file behind:

sudo rpm -ivh https://download.elasticsearch.org/elasticsearch/release/org/elasticsearch/distribution/rpm/elasticsearch/${elasticsearch_version}/elasticsearch-${elasticsearch_version}.rpm

---

You can also combine the download and extraction of kafka, so you can avoid cd-ing, avoid the temp file, and avoid running wget as root:

wget -qO - http://apache.mivzakim.net/kafka/${kafka_version}/kafka_${kafka_version_full}.tgz | sudo tar -zxvf - --directory=/usr/share

---

Also, consider installing kafka to /opt or /usr/local instead of /usr/share. /usr/share is typically for data, not programs. You can also flatten out the tarball a bit when you extract like so:

# Install kafka to /opt/kafka-0.9.0.1
wget -qO - ${url} | sudo tar -zxvf - --strip-components=1 --directory=/opt/kafka-${kafka_version}

---

wget options:

       -O file
       --output-document=file
           The documents will not be written to the appropriate files, but all will be concatenated together and written to file.  If - is used
           as file, documents will be printed to standard output, disabling link conversion.  (Use ./- to print to a file literally named -.)

tar options:

       -C, --directory=DIR
              change to directory DIR

       --strip-components=NUMBER
              strip NUMBER leading components from file names on extraction

You should use mktemp instead of creating your own temp file.

tempfile=$(sudo mktemp)  # sudo so the file will be owned by root, otherwise you'd have to `chown` it
echo -e "node.name: ${elasticsearch_nodename}\ncluster.name: ${elasticsearch_clustername}\nnetwork.bind_host: 0\n" | sudo tee ${tempfile} >/dev/null
cat /etc/elasticsearch/elasticsearch.yml | sudo tee --append ${tempfile} >/dev/null
sudo mv ${tempfile} /etc/elasticsearch/elasticsearch.yml

Better still would be to edit the YAML file in-place, with sed -i most likely (assuming the file already has node.name and friends already in there, but commented out or something).