Docker ElasticSearch S3 Snapshot

This is a docker image that has ElasticSearch load a specified snapshot from S3.

Usage

It requires these environmnent variables be set:

AWS_ACCESS_KEY_ID - Your S3 access key
AWS_SECRET_KEY - Your S3 secret key
SNAPSHOT_BUCKET - The S3 bucket that the snapshot is in
SNAPSHOT_ID - The snapshot you wish to restore from

Examples:

docker run -p 9200:9200 -e "AWS_ACCESS_KEY_ID=1234" \
                        -e "AWS_SECRET_KEY=1234" \
                        -e "SNAPSHOT_BUCKET=my_snapshots" \
                        -e "SNAPSHOT_ID=my_snapshot" \
                        heepster/docker-elasticsearch-s3-snapshot

Plugins

This image installs the elasticsearch-head plugin. You'll need to add the http.cors.enabled: true to the Elasticsearch configuration.

Dockerhub

This image is automatically built by Dockerhub. The repository is: https://hub.docker.com/r/heepster/docker-elasticsearch-s3-snapshot/

Configuration

Elasticsearch Config

You can pass additional elastic search config options by appending them to end of the docker run command:

-Des.node.name="TestNode"

Sane Health Check

Optionally, this docker offers the ability to give you back a sane health check while restoring the snapshot. If you query ElasticSearch's _cluster/health endpoint while restoring the snapshot, it will give you back something like

{ "status": "red",
...
}

The "red" is supposed to indicate that ElasticSearch isn't ready. However, the actual HTTP response is a 200, which throws off health check services that rely on a non-200 response to infer health status.

You can turn on the sane health check feature by setting an environment variable to true:

SANE_HEALTH_CHECK=true

Then, a new endpoint will give you either a 200 or 500 depending on whether or not the snapshot restoration is complete. That endpoint is:

<ip>:9400/

(Note that the port is different from ElasticSearch's default port.)

TODOs

If the Snapshot restoration process fails (e.g. if the snapshot doesn't exist), ElasticSearch's cluster health check will return a 'status: green' which means everything is working...Unfortunately, the sane health check feature will then relay this exactly. If the snapshot doesn't work, then I think the right thing to do is to kill the docker. Still need to think more about this.