Elastic Scaling in Spark 1.2 and Beyond


Details
Live Stream link: https://www.youtube.com/watch?v=lXgHut4QlNY
Spark 1.2 introduces dynamic allocation of cluster resources. This allows Spark applications to request more resources when the amount of work exceeds the capability of the existing set, and renounce resources when the opposite is true. In this talk the primary authors of this feature, Andrew Or and Aaron Davidson, will describe in detail how this feature works, the use cases it targets, and how to use it to maximize the utilization of resources in your Spark cluster. They will do live demonstration of its performance benefits and robustness to extreme failures.
The talk will be livestreamed, and the video will be published on the Apache Spark channel (https://www.youtube.com/user/TheApacheSpark) on YouTube.
Agenda
6:00pm - 7:00pm - Drinks, food, mingle
7:00pm - 8:30pm - Presentations
8:30pm - 9:00pm - More mingling
About the Host
GalvanizeU ( http://www.galvanizeu.com/ ) creates an innovation ecosystem designed to give entrepreneurs and innovators the best chance of success at the start of their next (or first) big thing. Through the three pillars of Capital, Community, and Curriculum, Galvanize builds a community greater than the sum of its parts to spark disruptive ideas and breakout companies.

Elastic Scaling in Spark 1.2 and Beyond