YARN is the next generation of Hadoop. It is a community effort, but mostly pushed forward by HortonWorks, Yahoo and other big companies. The YARN paper describes the architecture of YARN, and talks about its design rationale.
The paper won the best paper in SoCC’13 (Symposium of Cloud Computing). I put up some slides to discuss it in our group meeting. The paper itself is well-written and easy to understand, but I thought the slides might be interested to some one who wants to understand it quickly.
I used a day to make the slides, so it is by no means a refined work. You are welcome to adapt/enhance the slides however you like. [Apache Hadoop YARN]