Published on

Big data experiment with Elasticsearch, Logstash and Kibana


After hearing Michael Heap's talk on Endpoint 2014 about doing more useful stuff with your logfiles I got inspired to start experimenting with big data: Elastiscsearch, Logstash and Kibana.

I started the experiment by creating my source of "big data". As Michael had shown during his talk, I added some "info" level log lines to a backend application that logged certain events with some metadata to the log file.

I've found a very helpful guide to setup the ELK stack by DigitalOcean which I can wholeheartedly recommend. The only thing missing were the ansible scripts to provision your server quickly, which I found on valentinogagliardi's githubproject ansible-logstash

Once setup, the logs are collected through logstash, stored in elasticsearch and can get queried with a nice dashboard frontend with Kibana. It takes a while to get Kibana to display the information you want to see but it surely beats creating a custom dashboard from scratch. Look at those pretty graphs!

kibana in all its graphy glory