This template project provides the means to setup a locally running Elasticsearch-backed Kibana instance with zero configuration using Docker containers. The project also provides a minimalistic implementation of an indexer that is able to push data per-item or in-bulk into Elasticsearch (supports both synchronous and asynchronous programming models).
Docker as well as Docker Compose need to be installed on your system.
| Application | Version |
|---|---|
| Elasticsearch | 2.3.2 |
| Kibana | 4.5.0 |
Have a look at module elasticsearch-feeder. This module implements an example feeder in class com.mgu.analytics.feeder.Feeder. This class is derived from AbstractBaseFeeder and must implement the run method. The base class provides methods (cf. withAdapter) that create all contextual dependencies for you. So in your implementation of the run method, it suffices to do something like this:
@Override
public void run() {
withAdapter(search -> {
search.index(generateDocuments());
});
}withAdapter injects an already configured instance of ElasticAdapter (cf. module elasticsearch-adapter) into the method. The actual feeding logic is encoded using a ThrowingConsumer<T> and executed asynchronously in the background. The application waits for the termination of the ThrowingConsumer<T> (w/o timeouts).
Module elasticsearch-adapter provides a TypedDocument which is the abstract base class for any search document that you want to feed into Elasticsearch.
The resource file config.properties contains the configuration of the feeder. Currently, this only indicates the name of the index to use.
Change to the root directory of this project and issue the following command.
$ docker-compose upThis will fire up Elasticsearch and Kibana. After both applications launched successfully, you can access Kibana via your favourite browser at this URL.
This software is released under the terms of the MIT license.