February 14th, 2014


“The application is fast, reliable, and we could extend it, if necessary, so it’s perfect.”

– Enrico Aillaud, Co-Founder at Yameveo

Enrico Aillaud

Enrico Aillaud Co-Founder at Yameveo




Yameveo is a young company focused on the eCommerce field. We work mainly with Prestashop, Magento and Zend Framewok; we developed several modules, websites and we offer consultancies for technical matters like scalability problems and the creation of new features needed by our clients. We are collaborating with several clients located in Spain, Italy and the United States.


Solving scalability – from relational to Cassandra

We were contacted by a Spanish client who had scalability problems, the system they were working with wasn’t sufficiently powerful to serve in an efficient way all the customers requests on the web. The purpose of our project was to solve these problems related to the efficiency of the system so, we decided to migrate part of the information from a relational database to Cassandra.

At the very beginning of the project we thought about several relational solutions, like PostgreSQL, at that time we thought that it had several features which were perfect to finalise the project. Then, we obviously came at the question: “will this solve our scalability problems?”, the answer was no; for this reason we decided to focus on a NoSQL solution, we took advantage of the fact that our client was already using Cassandra for some parts of the business logic.

We have now been in production since march of 2013. The application is fast, reliable, and we could extend it, if necessary, so it’s perfect.


Running Cassandra

We are using an external datacenter. We have 5 servers with 32 GB of RAM, 8 CPU, 6 nodes in order to satisfy more or less 1 million visit per month and to store 500GB of data.

From the coding point of view, the application was made using Zend Framework and the library phpcassa in order to use Cassandra’s features in the framework.


Getting started

We found the documentation online quite exhaustive especially the one made by Datastax and we were able to develop the application and integrate Cassandra without any problem.