Apache Spark is a buzzing tool in the analytics world but you need the right platform to utilize its power. Docker with Mesos could be a great choice as the base of your analytics system that runs Spark. With this, you can follow the rapidly changing landscape and growing toolset. In a modern analytics platform, we have to deal with batch, interactive and streaming workload and this could be a challenge. The presentation outlines the technical aspects of our journey with Docker, Mesos and Spark. Our challenge was to support heterogeneous workload and running Spark alongside other applications.
Szuromi Tamás
Data scientist, Liligo
I’m Tamas Szuromi a versatile data scientist.
More generalist than specialist, information and data addict, developer, No&SQL specialist, statistics and machine learning enthusiast.
Every day I try to discover new technologies and methods to solve problems in the future. So I really like learning, reading, experienced by new technologies. I enjoy every field of Data Science where I can solve complex problems with data and the suitable tools or meanwhile just adding some useful values to the customer experience with insights or analysis.