caruccio/openshift-elasticsearch

OpenShift ElasticSearch Cartridge

Running scalable ElasticSearch on OpenShift.

The first gear is the master and data node. All other gears will play the data-only node role. In order to create data-only nodes, simply scale-up your app.

Usage

First of all, create your account on Getup Cloud: http://getupcloud.com. It is free and gives a 750h trial period.

Let's start by creating a new ES app called elastic:

$ rhc app create -s elastic https://reflector-getupcloud.getup.io/reflect?github=caruccio/openshift-elasticsearch
$ cd elastic

Point your browser to http://elastic-$namespace.getup.io to see it in action. You can contemplate your "cluster" in http://elastic-$namespace.getup.io/_cluster/nodes/stats?plugins=true. (Look mom! I've built a cluster of a single node!)

To add 4 more data-only ES instances, simply add more gears:

$ rhc cartridge-scale elasticsearch 4 -a elastic

Now you can be proud and show off with your amazing pile of pure distributed-real-time-analytics-multi-tenant-document-oriented-restful-schema-free-per-operation-persistent-open-source-o-matic server.

Plugins

The app is created with 3 installed plugins by default:

• mobz/elasticsearch-head
• karmi/elasticsearch-paramedic
• polyfractal/elasticsearch-inquisitor

In order to install/remove plugins, edit file plugins.txt, commit and push. If your plugin needs a full URL to download its code from, use NAME=URL to specify it. All plugins are installed on all gears.

Contributions

Please, anything...

Here are some suggestions:

• Download and install elasticsearch source on-the-fly
• User can choose what version it wants to run
• Better (?) cluster topology