Great review article:
“In this article I digested a number of MapReduce patterns and algorithms to give a systematic view of the different techniques that can be found in the web or scientific articles. Several practical case studies are also provided. All descriptions and code snippets use the standard Hadoopâs MapReduce model with Mappers, Reduces, Combiners, Partitioners, and sorting. This framework is depicted in the figure below.” (from Highly Scalable)
Worth reading if you are interested in how Hadoop and friends might apply to the problems that you are trying to solve: “MapReduce Patterns, Algorithms, and Use Cases”
Just one phrase that jumped out of the new missive from Zuck about privacy:
“As a matter of fact, privacy is so deeply embedded in all of the development we do that every day tens of thousands of servers worth of computational resources are consumed checking to make sure that on any webpage we serve, that you have access to see each of the sometimes hundreds or even thousands of individual pieces of information that come together to form a Facebook page.” (from Our Commitment to the Facebook Community)
When I tell (non computer) people that it takes hundreds or thousands of servers in hundreds of data centers to run something like Google or Facebook they are surprised.
Even I am surprised at “tens of thousands of servers worth of computational resources.”
Wait, why the nuanced phrasing? What is a server’s worth of computational resource? Is that a server or something else? Wait, what’s a server anyway? Oh well. Parsing.