Artikelformat

Review: Berlin Buzzwords 2014 – Searched, Stored and Scaled

Berlin Buzzwords

The fifth edition of Berlin Buzzwords was held from 25-28 May at the Kulturbrauerei and crealytics was a proud sponsor. My colleagues Andrei and Nicola as well as myself were enthusiastic participants and picked up a lot of technologies to explore and apply at our workplace. In short, it was a great experience with a really good mix of talks, a vibrant community and the possibility to meet people from all over the world.

The choice of Kulturbrauerei as a venue was an interesting one as we could feel the “buzz” as soon as we entered the gates. The former brewery has a distinctive architecture with many buildings, open areas and places to sit down and talk, eat or just chill out with friends.

Great talks throughout the whole conference

Things kicked off (officially) on 25th with a great keynote on Machine Learning applied in practice and the challenges of implementing a research heavy technology project in the real world. The advice offered by Ralf Herbrich was really useful. What was also interesting was the deep research and analysis that went behind the making of the TrueSkill ranking engine in the popular Xbox game Halo 3.

Buzzwords_1

The program was jam-packed with interesting sessions and it was hard to choose which track to attend. The organizers had done a good job to theme the sessions by Search / Store / Scale so that attendees could choose their track by the themes of their preference.

On the first day we did a mix and match of all the three tracks. It was great to learn about the capabilities of ElasticSearch and specially its query DSL, data aggregation and geo-spatial capabilities. Videos and slides for most of the talks are available here.

Another interesting technology which was presented was Apache Cassandra by the Datastax team. Cassandra has grown leaps and bound recently, and the performance and scalability features it offers are pretty impressive. With things like linear scaling, peer-to-peer architecture and guaranteed data safety it’s hard to overlook this awesome piece of technology.

The second day started with a bang with a great talk by Katherine Daniels on why devops need to be re-invented. If I remember one thing from the entire conference, it would be her talk. The situations and solutions described were spot-on and her delivery was fantastic.

We spent our second day learning about solving real world issues and scaling Solr / Lucene, the session by Dmitry Mescheryakov on using OpenStack to manage Hadoop clusters was great as well.

Buzzwords_2

Ted Dunning held an interesting talk about time series data bases. One of the most interesting parts of the talk was about building an anomaly detector. In this way you can visualize and understand your data better. Towards the end of the day, Itamar Syn-Hershko gave us an overview of the Elasticsearch plugins ecosystem. This was a practical and useful talk for programmers starting with Elasticsearch.

A conference worth participating

Overall we had a great time at the conference, the food was great and there were plenty of goodies given for free by the sponsors. It was good to meet some old friends and make some new ones as that is one great aspect of conferences all over the world. We hope to take part at the conference next year as well and are still buzzing with the knowledge and experience we gained.