At Populate, we use ElasticSearch as a superstructured data storage.

In the last days we have moved that infrastructure to Docker. This post describes how.

Creating the container:

# Create data volume
docker create -v /usr/share/elasticsearch/data/elasticsearch --name es-gobierto-budgets-data busybox bin/true

# Create elasticsearch
docker run -d  -p 9250:9200 --name es-gobierto-budgets --volumes-from es-gobierto-budgets-data elasticsearch

# Check it works
# using alpine
docker run -t -i --rm --volumes-from es-gobierto-budgets-data alpine /bin/sh
# using ubuntu
docker run -t -i --rm --volumes-from es-gobierto-budgets-data ubuntu /bin/bash

Backup the container:

docker run --rm --volumes-from es-gobierto-budgets-data -v $(pwd):/backup alpine tar cvf /backup/backup.tar /usr/share/elasticsearch/data/elasticsearch

Restore backup:

docker create -v /usr/share/elasticsearch/data/elasticsearch --name es-gobierto-budgets-data busybox bin/true
docker run --volumes-from es-gobierto-budgets-data -v $(pwd):/backup ubuntu bash -c "cd / && tar xvf /backup/es-gobierto-budgets-data.tar --strip 1 && mv share/elasticsearch/data/elasticsearch/* /usr/share/elasticsearch/data/elasticsearch/"