Setting up Elasticsearch on Debian - gnuhub/elasticsearch GitHub Wiki
These are alternative instructions for Debian Lenny, using system-directories.
Note: Elasticsearch will run as “root”
Check http://www.elasticsearch.com/download/ for the latest version of ElasticSearch.
curl -O -L https://github.com/downloads/elasticsearch/elasticsearch/elasticsearch-0.14.1.zip
unzip elasticsearch-* && rm elasticsearch-*.zip
mv elasticsearch-* /usr/local/elasticsearchCreate a file “/etc/init.d/elasticsearch” with the following content:
#! /bin/sh
### BEGIN INIT INFO
# Provides: elasticsearch
# Required-Start: $all
# Required-Stop: $all
# Default-Start: 2 3 4 5
# Default-Stop: 0 1 6
# Short-Description: Starts elasticsearch
# Description: Starts elasticsearch using start-stop-daemon
### END INIT INFO
ES_HOME=/usr/local/elasticsearch
ES_MIN_MEM=256m
ES_MAX_MEM=2g
DAEMON=$ES_HOME/bin/elasticsearch
NAME=elasticsearch
DESC=elasticsearch
PID_FILE=/var/run/$NAME.pid
LOG_DIR=/var/log/$NAME
DATA_DIR=/var/lib/$NAME
WORK_DIR=/tmp/$NAME
CONFIG_FILE=/etc/$NAME/elasticsearch.yml
DAEMON_OPTS="-p $PID_FILE -Des.config=$CONFIG_FILE -Des.path.home=$ES_HOME -Des.path.logs=$LOG_DIR -Des.path.data=$DATA_DIR -Des.path.work=$WORK_DIR"
test -x $DAEMON || exit 0
set -e
case "$1" in
start)
echo -n "Starting $DESC: "
mkdir -p $LOG_DIR $DATA_DIR $WORK_DIR
if start-stop-daemon --start --pidfile $PID_FILE --startas $DAEMON -- $DAEMON_OPTS
then
echo "started."
else
echo "failed."
fi
;;
stop)
echo -n "Stopping $DESC: "
if start-stop-daemon --stop --pidfile $PID_FILE
then
echo "stopped."
else
echo "failed."
fi
;;
restart|force-reload)
${0} stop
sleep 0.5
${0} start
;;
*)
N=/etc/init.d/$NAME
echo "Usage: $N {start|stop|restart|force-reload}" >&2
exit 1
;;
esac
exit 0Install the script:
chmod +x /etc/init.d/elasticsearch
update-rc.d elasticsearch defaultsCreate /etc/elasticsearch/elasticsearch.yml and configure to your needs.
A basic example:
cluster:
name: MyCluster
network:
host: _eth0_/etc/init.d/elasticsearch start
/etc/init.d/elasticsearch stop