Data Index inside the Cloud: How to Deploy Elasticsearch at Jelastic

| March 3, 2016

image001Elasticsearch is a powerful open source engine for search and analytics, that makes data easy to store and explore. Being focused on reliability, real-time analytics and simple management, it is based on high performance and full-featured Java-powered Apache Lucene library, which helps to implement the provided advanced search and analytics capabilities. Herewith, a special RESTful API allows to perform any action by means of a simple JSON over HTTP, making this solution truly developer-friendly.

Since Elasticsearch engine is rather popular and highly demanded among our users, we’ve decided to describe how it can be deployed in a matter of minutes at any of our partner’s Platforms, leveraging tight integration of Docker containers support at Jelastic Cloud.

So, below we’ll show you how to set up your own Elasticsearch server using the appropriate Docker template, launch it, and even how to move to another app version (e.g. upon the new one is released). Let’s consider all of these points one-by-one.

Deploy Elasticsearch Docker Container

Being based on Docker standard, the Elasticsearch image can be provisioned for production-ready use at Jelastic PaaS in a few clicks, just like it’s described below:

1. Log into your Jelastic dashboard and click the New environment button at the top pane. In the opened wizard, switch to the Docker tab and click Select Container in the middle of the frame.image12

2. At the Search tab, type elasticsearch into the corresponding field and press Enter. Choose the same-named image within the appeared list by simply clicking on it.image07

Then, select the version you’d like to install by means of the drop-down menu at the top (as an example, we’ll choose the 2.0.2 one) and click Next to proceed.

3. Now, you can configure the general parameters of your environment - adjust the amount of allocated resources using cloudlet sliders, select the desired region and type environment name image10

Note: Elasticsearch requires either a Public IP address to be attached to a container (for that, use the switcher circled in the image above) or have the appropriate endpoints set in order to work properly. Herewith, the first option is paid but more efficient, whilst endpoints can be used free of charge (we’ll show how to appoint them right down this article).

To submit the creation, click the same-named button when ready.

4. That's all! In a minute, you'll get your ready-to-use Docker-based environment with Elasticsearch already deployed inside. image11

Launch Elasticsearch

Now, let’s ensure our Elasticsearch application is accessible and works fine - this can be confirmed through receiving a piece of JSON upon accessing its start page. As it was mentioned above, depending on your preferences you may connect via:

  • external IP address - the common choice, which ensures simple instance handling and faster response time, whilst being a paid option
  • endpoints - the free of charge alternative, provided by Jelastic

Connection via Public IP

In case you’ve attached an external IP address to your elasticsearch container, simply click the Open in browser button next to your environment and add the default 9200 port at the end of the address within the opened tab:image02

This way is standard and, as you can see, is maximally optimized for convenient usage.

Connection via Endpoints

If you’d like to leverage endpoints solution instead of spending the additional costs for external address utilization, you need to take into consideration that Elasticsearch requires the http and transport.tcp services to be accessible from outside. Here, the first service is used for the incoming HTTP requests’ processing, and the second one - for communication between nodes.

So, to make your deployed search engine to work properly, you’ll need to manually expose the appropriate ports via Jelastic Shared Load-Balancer for the corresponding container.

1. For that, click the Settings icon next to your environment, switch to the Endpoints section and select Add, like it is shown in the image below:image05

2. Within the opened frame, define the required parameters for the first endpoint (i.e. the http service entry point), exposing the 9200 container port:image01

Click Add to create.

3. Repeat the previous step for the transport service (9300 port). As a result, you should receive two endpoints, similar to the shown below:image06

For now, you are able to operate with your elasticsearch instance using the links within the Access URL column. To ensure everything is fine, navigate to the http one:image04

That’s it! Now you can start building your own index of the required document-oriented data.

Update Elasticsearch to New Version

Jelastic provides an easy way to switch between the available Docker image tags, whilst saving all the inherent server configurations and data stored, with the distinctive container redeploy feature. So, in case of the necessity to upgrade your Elasticsearch instance to the newly released version, just follow the next simple steps:

1. Click Redeploy container next to the required Docker container (or, if operating with several nodes, select the same button for the corresponding image name to update all of them at once).image09

2. Within the appeared frame, choose the desired Tag your application should be updated with (e.g. the latest one) and click Redeploy.image08

Be aware that the operation of redeployment will cause a temporary downtime of your application. However, this can be easily avoided with Sequential deployment of multiple image instances.

3. In a minute or so, you'll get your container updated. To check the exact Elasticsearch version that is currently run, just refer to its start page.image03

In such a way, you can move to the newest application version right after it is released without the mess of complex adjustments and special configurations.

Have met any problems with installation? Or would like to request the instruction for any other application? Just let us know within the comments below or at Stackoverflow.

Hope you’ve enjoyed our guide! And remember that you always can try it by yourself through creating your own trial account completely for free.