In this example, we will discuss about How to create a Spring Boot, Spring Data and Elasticsearch Example.
- Spring Boot
- Spring Boot Starter Data Elasticsearch
- Elasticsearch 2.4
- Kibana 4.6
SpringBoot 1.5.*.RELEASE supports only ElasticSearch 2.4.0. They don’t support the latest version of ElasticSearch 5.x version. Read this – Spring Data Elasticsearch Spring Boot version matrix and ELK Compatibility
Run this command in the docker container to install elastic search
docker run -d -it --name kuldeep -p 9200:9200 -p 9300:9300 -e ES_JAVA_OPTS="-Xms1g -Xmx1g" -m 1500m elasticsearch:2.4
Run this command in the docker container to install kibana
docker run -d -it --name kibanak --link kuldeep:elasticsearch -p 5601:5601 kibana:4.6
Checking with docker ps command, all their container should be running
-
Download the code and import it to Eclipse or any Id
-
Run the project and make sure Java Container is able to connect Elastic Search
-
Call the SaveContact URL (http://localhost:8080/user/savecontact) to create a record
POST /user/savecontact HTTP/1.1 Host: localhost:8080 Content-Type: application/json Cache-Control: no-cache Postman-Token: 1d8e66f0-ea57-5ea5-4b6a-d9be1c751812
{ "id":"4", "contactId":"104", "name": "hello4", "email":"hello4@123.com", "company":"xcompany", "contactType":"3" }
- Access the Kibana URL http://192.168.99.100:5601/app/kibana and go to Settings -> Indices
- Next Step is to create Index, type contacts in the textbox and uncheck the "Index contains time based events"
- Once Index is created we can check all the fields mentioned in bean has been created
- Finally we can go to discover tab to check the data
We can do all the CRUD Operation here.