Step‑by‑Step Guide to Install an ElasticSearch 7.17.x Cluster on Ubuntu

Overview

The article explains how to build a three‑node ElasticSearch 7.17.x cluster on Ubuntu 22.04 using binary packages. It covers prerequisite software, host and network configuration, system tuning, user and directory setup, ElasticSearch installation, service definition, startup, and health verification.

Environment Information

The cluster consists of three servers:

ELK01 : IP 10.0.0.40, Ubuntu 22.04

ELK02 : IP 10.0.0.41, Ubuntu 22.04

ELK03 : IP 10.0.0.42, Ubuntu 22.04

Practical Steps

Install JDK (required on all nodes)

ElasticSearch runs on Java, so JDK 11 is installed on each host.

# wget https://mirrors.huaweicloud.com/openjdk/11.0.2/openjdk-11.0.2_linux-x64_bin.tar.gz
# tar -xvf openjdk-11.0.2_linux-x64_bin.tar.gz
# ln -s /root/jdk-11.0.2 /usr/local/jdk11
# vim /etc/profile   # add:
#   export JAVA_HOME=/usr/local/jdk11
#   export PATH=$JAVA_HOME/bin:$PATH
# source /etc/profile
# java -version   # should show OpenJDK 11.0.2

Configure Hostnames and /etc/hosts

Set each node's hostname and add name‑to‑IP mappings.

# hostnamectl set-hostname ELK01   # on ELK01
# hostnamectl set-hostname ELK02   # on ELK02
# hostnamectl set-hostname ELK03   # on ELK03

# vim /etc/hosts
10.0.0.40 ELK01
10.0.0.41 ELK02
10.0.0.42 ELK03

Synchronize System Time

Install ntpdate and schedule periodic sync.

# ln -svf /usr/share/zoneinfo/Asia/Shanghai /etc/localtime
# apt -y install ntpdate
# ntpdate ntp.aliyun.com
# echo "*/5 * * * * /usr/sbin/ntpdate ntp.aliyun.com" > /var/spool/cron/crontabs/root

System Parameter Tuning (all nodes)

Increase file descriptors, virtual memory, and network limits for ElasticSearch.

# vim /etc/sysctl.conf
fs.file-max = 655360
vm.max_map_count = 2147483642
vm.swappiness = 1
net.ipv4.tcp_keepalive_time = 600
net.ipv4.tcp_keepalive_intvl = 60
net.ipv4.tcp_keepalive_probes = 10
net.ipv4.tcp_max_syn_backlog = 4096
net.core.somaxconn = 4096
net.core.netdev_max_backlog = 16384
net.core.rmem_max = 262144
net.core.wmem_max = 262144
# sysctl -p
# sysctl -q vm.max_map_count   # verify value

Create Directories and Elasticsearch User

Prepare data and log directories and a dedicated user.

# mkdir -p /data/elasticsearch
# mkdir -p /var/log/elasticsearch
# useradd elasticsearch
# chown elasticsearch:elasticsearch -R /data/elasticsearch
# chown elasticsearch:elasticsearch -R /var/log/elasticsearch

Set User Limits

Allow the elasticsearch user to open many files and processes.

# vim /etc/security/limits.conf
elasticsearch   hard    nofile  655360
elasticsearch   soft    nofile  655360
elasticsearch   hard    nproc   8192
elasticsearch   soft    nproc   8192
elasticsearch   hard    memlock unlimited
elasticsearch   soft    memlock unlimited

Download and Extract ElasticSearch

Obtain the 7.17.x tarball from the official archive, extract it, and create a symlink.

# wget https://artifacts.elastic.co/downloads/elasticsearch/elasticsearch-7.17.26-linux-x86_64.tar.gz
# tar -xvf elasticsearch-7.17.26-linux-x86_64.tar.gz
# mv elasticsearch-7.17.26 /data/
# chown elasticsearch:elasticsearch -R /data/elasticsearch-7.17.26
# ln -s /data/elasticsearch-7.17.26 /usr/local/es7

Configure ElasticSearch

Edit /usr/local/es7/config/elasticsearch.yml with cluster settings.

# vim /usr/local/es7/config/elasticsearch.yml
cluster.name: es7
path.data: /data/elasticsearch
path.logs: /var/log/elasticsearch
network.host: 0.0.0.0
http.port: 9200
discovery.seed_hosts: ["ELK01", "ELK02", "ELK03"]
cluster.initial_master_nodes: ["ELK01", "ELK02", "ELK03"]
# node.name can be set per host, e.g. node.name: ELK01

Create systemd Service

Define a service unit to manage ElasticSearch.

# vim /lib/systemd/system/es.service
[Unit]
Description=elasticsearch service
Documentation=https://www.elastic.co/guide/en/elasticsearch/reference/7.17/index.html
After=network.target auditd.service

[Service]
LimitMEMLOCK=infinity
User=elasticsearch
ExecStart=/usr/local/es7/bin/elasticsearch
TimeoutStopSec=0
TimeoutStartSec=0

[Install]
WantedBy=multi-user.target

Start the Cluster

Reload systemd, enable and start the service on each node.

# systemctl daemon-reload
# systemctl start es
# systemctl status es   # should show active (running)

Verify Cluster Health

Check that all three nodes are joined and the cluster is green.

# curl 10.0.0.40:9200/_cat/nodes
# curl 10.0.0.40:9200/_cat/health?v

When the health output shows green and lists three nodes, the ElasticSearch cluster is successfully deployed.