#3 => David Greenberg on Sparrow: Distributed, Low Latency Scheduling

We're thrilled to have David Greenberg (https://github.com/dgrnbrg) presenting Sparrow: Distributed, Low Latency Scheduling (http://people.csail.mit.edu/matei/papers/2013/sosp_sparrow.pdf) by Kay Ousterhout, Patrick Wendell, Matei Zaharia and Ion Stoica.

Intro

When you need to execute code on a cluster of machines, deciding which machine should run that code becomes a complex problem, known as scheduling. We're all familiar with routing problems, such as the recent RapGenius incident (http://news.rapgenius.com/James-somers-herokus-ugly-secret-annotated). It turns out that simple improvements to randomized routing can dramatically improve the performance! Sparrow is a distributed scheduling algorithm for low latency, high throughput workloads. We'll review the Sparrow algorithm, and learn the tricks that they used. Then, we'll discuss other applications of Sparrow, besides the big-data map-reduce application it was created for.

Bio

David Greenberg loves learning. He currently works at Two Sigma, where he leads several efforts to build reliable, distributed cluster computing environments. He enjoys reading a good paper with glass of scotch on a cold winter night. Given this winter, there's been a lot of papers :)

http://photos3.meetupstatic.com/photos/event/1/b/8/c/600_348067052.jpeg

Details

Doors open at 7 pm; the presentation will begin at 7:30 pm; and, yes, there will be beer and pizza.

After David presents the paper, we will open up the floor to discussion and questions.

We hope that you'll read the paper before the meetup, but don't stress if you never get around to it, and if you have any questions, thoughts, or related information, please visit our github-thread (https://github.com/papers-we-love/papers-we-love/issues/89) on the matter.

Additionally, if you have any papers you want to add to the repository above, please send us a pull request. Also, if you have any ideas/questions about this meetup or the Papers-We-Love org, just open up an issue.