Summary: Introduction to Graph Databases and Neo4j (28/11)

Hi devs! How's the new year going? We hope everything goes well ;-)

As you remember, last 28th of November we received the visit of Neotechnology, the creators of Neo4j and they gave us a talk about this NoSQL database plus an introduction to Graph Databases in our community. Here is the summary of this talk by Stefan Armbrustrer, and of course, we would like to thank Stefan and Dirk for coming from Munich to talk about Neo4J inside our community, it was really a pleasure to host their talk.

First Stefan started his talk speaking about the trends and what is all about with terms like BigData and NoSQL: we need to rethink about how the data is increasing in size really really fast, why everything is almost connected in our IT world and how is increasingly connected (graphs are everywhere). 

Next, the talk continued with the description about what is NoSQL and why is not the magic key to solve all the problems related with query and save data. First, in the process of comparing a relational database with a non-relational one, we must think of what kind of NoSQL database to choose, and we have a bunch of different types: key/value, column-oriented, document, graph, etc. Depending on which are the problems you are trying to solve, the nature of your database will fit better one or another (ex: cost of the size vs complexity of your database).

When you think in a graph database, you must rethink your model in a way of connected nodes with relationships that have properties (key+value). Of course you can use indexes and in the case of Neo4j another functionalities like labels. Using the example of a social network, Stefan presented numbers comparing the performance of Neo4J vs another relational database and the results were really impressive: the rdbms was really slow (1000x) and increasing the number of records to a million didn't change the time response of the query, in fact, that's the key of the power of a graph database.

In this sense, Stefan showed us the strengths of a graph database for example, how powerful the model you can build like your can do in relational database, or how fast it is and how easy is to query, as well as their constrains, including the learning curve or your conceptual change you must do. The presentation went on the benefits of using Neo4j, and how powerful and functionalities it has: it can be embedded and use it with a lot of languages, provides acid transactions, integrates indexes and a REST API, high-availability, highly scalable, etc. Below, Stefan showed us in live how to play with Neo4J, and how beautiful is to query using a browser or with Java code and how easy is to work with this schema-free database. 

The talk ended with the sample of many of the customers Neotechnology are working with, and how they are satisfied using Neo4j and at the end, the closure was the raffle with few books about Graph databases that helped to end with a better taste this great talk. 

Ah! Here you can find the video in our Youtube channel as well as the presentation in our Slideshare account. Hope you will enjoy it as we did ;-)

Finally, from the BarcelonaJUG crew, we would like the best for this new year ...and happy coding!