ElasticSearch for Docker

This is a highly configurable ElasticSearch Docker image built using Docker's automated build process published to the public Docker Hub Registry.

It is usually the back-end for a Logstash instance with Kibana as the frontend forming what is commonly referred to as an ELK stack.

How to use this image

To start a basic container using ephemeral storage, exposing port 9200 for client connectivity:

docker run --name elasticsearch \
  --publish 9200:9200 \
  cgswong/elasticsearch

Within the container the volume /var/lib/elasticsearch is exposed. It contains the sub-directories for data, log and config. To start a default container with attached persistent/shared storage for data:

mkdir -p /es/data
docker run --name elasticsearch
  --publish 9200:9200 \
  --volume /data:/var/lib/elasticsearch/data \
  cgswong/elasticsearch

Attaching persistent storage ensures that the data is retained across container restarts (with some obvious caveats). It is recommended this be done instead via a data container, preferably hosted an AWS S3 bucket or other externalized, distributed persistent storage.

Available Features

A few plugins are installed:

• BigDesk: Provides live charts and statistics for an Elasticsearch cluster. You can open a browser and navigate to http://localhost:9200/_plugin/bigdesk/ it will open Bigdesk and auto-connect to the ES node. You will need to change the localhost and 9200 port to the correct values for your environment/setup.

• ElasticHQ: Monitoring, management, and querying web interface for ElasticSearch instances and clusters.

• Whatson: an ElasticSearch plugin to visualize the state of a cluster.

• Kopf: A simple web administration tool for Elasticsearch.

• AWS Cloud - Allows usage of AWS API for unicast discovery and S3 repositories for snapshots.

Any commands passed on the command line are accepted as input to the elasticsearch command. For example, to set the JVM heap size to 10GB:

docker run --name elasticsearch \
  --publish 9200:9200 \
  cgswong/elasticsearch -Xmx10g -Xms10g

Additional Configuration

Environment variables are accepted as a means to provide further configuration by reading those starting with ES_. Any matching variables will get added to ElasticSearch's configuration file, `elasticsearch.yml' by:

1. Removing the ES_ prefix
2. Transforming to lower case
3. Replacing occurrences of _ with ., except where there is a double (__) which is replaced by a single (_).

For example, an environment variable ES_CLUSTER_NAME=lscluster will result in cluster.name=lscluster within elasticsearch.yml. Similarly, ES_CLOUD_AWS_ACCESS__KEY=GHKDFIADFNADFIADFKJG would result in cloud.aws.access_key=GHKDFIADFNADFIADFKJG within elasticsearch.yml.

You can also import your own configuration file by setting ES_CFG_URL to a valid URL. The environment variable substitution can then also be used for the appropriate variables within your file as well.

As an example:

docker run --name elasticsearch \
  --publish 9200:9200 \
  --env ES_CFG_URL=http://[some url] \
  --env ES_CLUSTER_NAME=es_test_01 \
  cgswong/elasticsearch

Exposed Ports

• 9200/tcp: For client connectivity.
• 9300/tcp: For clustering using ransport/node protocols.