Monday, October 7, 2013

Benchmarking Graph Databases with relational databases


Graph data management has recently received a lot of attention, particularly with the explosion of social media and other complex, inter-dependent datasets. As a result, a number of graph data management systems have been proposed. But this brings us to the question: What happens to the good old relational database systems (RDBMSs) in the context of graph data management?

Do RDBMSs become obsolete or are they still relevant for large-scale graph data management? Do users now need to dump their data from the relational database to a graph database, or can they store and query graphs (with comparable or better performance) from within the relational engine?

Unfortunately, there’s an absence of a comparative study of modern relational databases in the context of graph data management. This is the problem we hope to solve at the MIT CSAIL Database Group with our study to benchmark and understand the performance of relational databases over very-large graphs. Our early results are promising.

Read more here

Updated : 15th Oct 2013

After our last post about benchmarking graph databases, Neo Technology representatives contacted us and said that they repeated the shortest path queries on Neo4j with the Facebook dataset and they observed 2-4 orders of magnitude better performance. This was an intriguing comment and we were very interested in finding out why this could be the case.

Read about the updated results here

Leave a Reply

All Tech News IN © 2011 & Main Blogger .