This first series of posts presents my home mini-cluster that I have set up in September 2015.
Having only a Windows laptop, I had decided to buy a cheap SBC (Single Board Computer), sometimes called SoC (System on a Chip), to have my own Linux server at home.
I decided to buy a MinnowBoard MAX (choice explained in Part I) and was very satisfied with it. It has enough specs for any home usage, however it was not very powerful to make calculations, and was slow to process a lot of data compared to my laptop. So I thought why not buy a few more and make a cluster !
But can I actually make a powerful cluster with only SBCs ? How fast can it be ? How many nodes does it take to be faster than a single SBC ? Can it be faster than my Core-i7 laptop ?
I explain in the different posts of this series how I selected the different components (Part I) and how I set up the cluster (Part II). Then I show how I installed popular distributed computing tools, Hadoop and Spark (Part III), in order to finally benchmark a computing task (Part IV) on different platforms, which will give us the answers to these questions.
Here are the different parts :