GoogleCloudPlatform/elasticsearch-docker

elasticsearch-docker

Dockerfile source for elasticsearch docker image.

Upstream

This source repo was originally copied from: https://github.com/docker-library/elasticsearch

Disclaimer

This is not an official Google product.

About

This image contains an installation Elasticsearch

For more information, see the Official Image Marketplace Page.

Pull command (first install gcloud):

gcloud auth configure-docker && docker -- pull marketplace.gcr.io/google/elasticsearch7

Dockerfile for this image can be found here

Table of Contents

• Using Kubernetes
  • Run Elasticsearch
    • Start an Elasticsearch instance
    • Use a persistent data volume
  • Using Elasticsearch
    • Connect and start using Elasticsearch
  • Configurations
    • Using configuration volume
• Using Docker
  • Run Elasticsearch
    • Start an Elasticsearch instance
    • Use a persistent data volume
  • Using Elasticsearch
    • Connect and start using Elasticsearch
  • Configurations
    • Using configuration volume
  • Clustering
    • Creating simple cluster
• References
  • Ports
  • Volumes

Using Kubernetes

Consult Marketplace container documentation for additional information about setting up your Kubernetes environment.

Run Elasticsearch

Start an Elasticsearch instance

Copy the following content to pod.yaml file, and run kubectl create -f pod.yaml.

apiVersion: v1
kind: Pod
metadata:
  name: some-elasticsearch
  labels:
    name: some-elasticsearch
spec:
  containers:
    - image: marketplace.gcr.io/google/elasticsearch7
      name: elasticsearch

Run the following to expose the port. Depending on your cluster setup, this might expose your service to the Internet with an external IP address. For more information, consult Kubernetes documentation.

kubectl expose pod some-elasticsearch --name some-elasticsearch-9200 \
  --type LoadBalancer --port 9200 --protocol TCP

Elasticsearch host requires configured enviroment. On Linux host please run sysctl -w vm.max_map_count=262144. For details please check official documentation.

To retain Elasticsearch data across container restarts, see Use a persistent data volume.

To configure your application, see Configurations.

Use a persistent data volume

To retain Elasticsearch data across container restarts, we should use a persistent volume for /usr/share/elasticsearch/data.

Copy the following content to pod.yaml file, and run kubectl create -f pod.yaml.

apiVersion: v1
kind: Pod
metadata:
  name: some-elasticsearch
  labels:
    name: some-elasticsearch
spec:
  containers:
    - image: marketplace.gcr.io/google/elasticsearch7
      name: elasticsearch
      volumeMounts:
        - name: elasticsearchdata
          mountPath: /usr/share/elasticsearch/data
  volumes:
    - name: elasticsearchdata
      persistentVolumeClaim:
        claimName: elasticsearchdata
---
# Request a persistent volume from the cluster using a Persistent Volume Claim.
kind: PersistentVolumeClaim
apiVersion: v1
metadata:
  name: elasticsearchdata
  annotations:
    volume.alpha.kubernetes.io/storage-class: default
spec:
  accessModes: [ReadWriteOnce]
  resources:
    requests:
      storage: 5Gi

Run the following to expose the port. Depending on your cluster setup, this might expose your service to the Internet with an external IP address. For more information, consult Kubernetes documentation.

kubectl expose pod some-elasticsearch --name some-elasticsearch-9200 \
  --type LoadBalancer --port 9200 --protocol TCP

Using Elasticsearch

Connect and start using Elasticsearch

Attach to the container.

kubectl exec -it some-elasticsearch -- bash

The following examples use curl. First we need to install it as it is not installed by default.

apt-get update && apt-get install -y curl

We can get test data into Elasticsearch using a HTTP PUT request. This will populate Elasticsearch with test data.

curl -H "Content-Type: application/json" -XPUT http://localhost:9200/estest/test/1 -d \
'{
   "name" : "Elasticsearch Test",
   "Description": "This is just a test"
 }'

We can try searching for our test data using curl.

curl http://localhost:9200/estest/_search?q=Test

Configurations

Using configuration volume

Assume /path/to/your/elasticsearch.yml is the configuration file on your localhost. We can mount this as volume at /usr/share/elasticsearch/config/elasticsearch.yml on the container for Elasticsearch to read from.

Create the following configmap:

kubectl create configmap elasticsearchconfig \
  --from-file=/path/to/your/elasticsearch.yml

Copy the following content to pod.yaml file, and run kubectl create -f pod.yaml.

apiVersion: v1
kind: Pod
metadata:
  name: some-elasticsearch
  labels:
    name: some-elasticsearch
spec:
  containers:
    - image: marketplace.gcr.io/google/elasticsearch7
      name: elasticsearch
      volumeMounts:
        - name: elasticsearchconfig
          mountPath: /usr/share/elasticsearch/config
  volumes:
    - name: elasticsearchconfig
      configMap:
        name: elasticsearchconfig

Run the following to expose the port. Depending on your cluster setup, this might expose your service to the Internet with an external IP address. For more information, consult Kubernetes documentation.

kubectl expose pod some-elasticsearch --name some-elasticsearch-9200 \
  --type LoadBalancer --port 9200 --protocol TCP

See Elasticsearch documentation on available configuration options.

Also see Volume reference.

Using Docker

Consult Marketplace container documentation for additional information about setting up your Docker environment.

Run Elasticsearch

Start an Elasticsearch instance

Use the following content for the docker-compose.yml file, then run docker-compose up.

version: '2'
services:
  elasticsearch:
    container_name: some-elasticsearch
    image: marketplace.gcr.io/google/elasticsearch7
    ports:
      - '9200:9200'

Or you can use docker run directly:

docker run \
  --name some-elasticsearch \
  -p 9200:9200 \
  -d \
  marketplace.gcr.io/google/elasticsearch7

Elasticsearch host requires configured enviroment. On Linux host please run sysctl -w vm.max_map_count=262144. For details please check official documentation.

To retain Elasticsearch data across container restarts, see Use a persistent data volume.

To configure your application, see Configurations.

Use a persistent data volume

To retain Elasticsearch data across container restarts, we should use a persistent volume for /usr/share/elasticsearch/data.

Assume /path/to/your/elasticsearch/data is a persistent data folder on your host.

Use the following content for the docker-compose.yml file, then run docker-compose up.

version: '2'
services:
  elasticsearch:
    container_name: some-elasticsearch
    image: marketplace.gcr.io/google/elasticsearch7
    ports:
      - '9200:9200'
    volumes:
      - /path/to/your/elasticsearch/data:/usr/share/elasticsearch/data

Or you can use docker run directly:

docker run \
  --name some-elasticsearch \
  -p 9200:9200 \
  -v /path/to/your/elasticsearch/data:/usr/share/elasticsearch/data \
  -d \
  marketplace.gcr.io/google/elasticsearch6

Using Elasticsearch

Connect and start using Elasticsearch

Attach to the container.

docker exec -it some-elasticsearch bash

The following examples use curl. First we need to install it as it is not installed by default.

apt-get update && apt-get install -y curl

We can get test data into Elasticsearch using a HTTP PUT request. This will populate Elasticsearch with test data.

curl -H "Content-Type: application/json" -XPUT http://localhost:9200/estest/test/1 -d \
'{
   "name" : "Elasticsearch Test",
   "Description": "This is just a test"
 }'

We can try searching for our test data using curl.

curl http://localhost:9200/estest/_search?q=Test

Configurations

Using configuration volume

Assume /path/to/your/elasticsearch.yml is the configuration file on your localhost. We can mount this as volume at /usr/share/elasticsearch/config/elasticsearch.yml on the container for Elasticsearch to read from.

Use the following content for the docker-compose.yml file, then run docker-compose up.

version: '2'
services:
  elasticsearch:
    container_name: some-elasticsearch
    image: marketplace.gcr.io/google/elasticsearch7
    ports:
      - '9200:9200'
    volumes:
      - /path/to/your/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml

Or you can use docker run directly:

docker run \
  --name some-elasticsearch \
  -p 9200:9200 \
  -v /path/to/your/elasticsearch.yml:/usr/share/elasticsearch/config/elasticsearch.yml \
  -d \
  marketplace.gcr.io/google/elasticsearch7

See Elasticsearch documentation on available configuration options.

Also see Volume reference.

Clustering

Creating simple cluster

In the following guide you will learn how to create simple two-node Elasticsearch cluster. This is only an example how to configure and link together containers, not a production ready configuration. For production ready configuration please refer to official documentation.

We will need a master node, that will also serve as a gateway to our cluster. Single agent node will be attached to the master node.

Use the following content for the docker-compose.yml file, then run docker-compose up.

version: '2'
services:
  elasticsearch-master:
    container_name: some-elasticsearch-master
    image: marketplace.gcr.io/google/elasticsearch7
    ports:
      - '9200:9200'
    command:
      - '-Enetwork.host=0.0.0.0'
      - '-Etransport.tcp.port=9300'
      - '-Ehttp.port=9200'
  elasticsearch-agent:
    container_name: some-elasticsearch-agent
    image: marketplace.gcr.io/google/elasticsearch7
    command:
      - '-Enetwork.host=0.0.0.0'
      - '-Etransport.tcp.port=9300'
      - '-Ehttp.port=9200'
      - '-Ediscovery.zen.ping.unicast.hosts=some-elasticsearch-master'
    depends_on:
      - elasticsearch-master

Or you can use docker run directly:

# elasticsearch-master
docker run \
  --name some-elasticsearch-master \
  -p 9200:9200 \
  -d \
  marketplace.gcr.io/google/elasticsearch7 \
  -Enetwork.host=0.0.0.0 \
  -Etransport.tcp.port=9300 \
  -Ehttp.port=9200

# elasticsearch-agent
docker run \
  --name some-elasticsearch-agent \
  --link some-elasticsearch-master \
  -d \
  marketplace.gcr.io/google/elasticsearch7 \
  -Enetwork.host=0.0.0.0 \
  -Etransport.tcp.port=9300 \
  -Ehttp.port=9200 \
  -Ediscovery.zen.ping.unicast.hosts=some-elasticsearch-master

After few seconds, we can check that cluster is running invoking http://localhost:9200/_cluster/health.

References

Ports

These are the ports exposed by the container image.

Port	Description
TCP 9114	Prometheus exporter port
TCP 9200	Elasticsearch HTTP port.
TCP 9300	Elasticsearch default communication port.

Volumes

These are the filesystem paths used by the container image.

Path	Description
/usr/share/elasticsearch/data	Stores Elasticsearch data.
/usr/share/elasticsearch/config/elasticsearch.yml	Stores configurations.
/usr/share/elasticsearch/config/log4j2.properties	Stores logging configurations.