This docker-compose project creates a Elasticsearch cluster with the following topology:
- 5 data nodes
- 3 dedicated master nodes
- 1 dedicated client node
- 1 kibana node pointing to the client node
- Timelion plugin installed
- Sense plugin installed
- 1 dockerbeat node submitting container metrics to Elasticsearch
It is very convenient for those situation where you need a more realistic ES deployment than just a couple of nodes running on localhost.
- docker 1.9
- docker-compose 1.6
To start the cluster:
> docker-compose up
Kibana will be available in the following URL:
http://<your_docker_host>:5601
Cluster endpoint via client node:
> curl -XGET <your_docker_host>:9200
For example to check the nodes in the cluster:
> curl -XGET <your_docker_host>/_nodes?pretty=true
To visualize the Dockerbeat dashboard follow the following steps:
- Configure an index pattern with the pattern 'dockerbeat-*'
- Import visualizations and dashboard in kibana-dashboards/dockerbeat.json
- Enjoy