Concurrent Futures in Sheepdoge: How a few lines of code resulted in a 78% performance improvement

Concurrent Futures in Sheepdoge: How a few lines of code resulted in a 78% performance improvement

For the past couple of months, I’ve been working on Sheepdoge, a tool for managing your personal Unix machines with Ansible. It’s like boxen, but for Ansible. One new sheepdoge feature I’m particularly excited about is the use of concurrent.futures during sheepdoge install. concurrent.futures provides a high-level API for executing code asynchronously, making adding thread/process based concurrency trivial. It is a standard library python module as of 3.2, and is available on python 2.

Predictive Auto-scaling in Kubernetes - Thesis

Over the course of my final year at Williams College, I spent a lot of time working on a distributed systems thesis with Professor Jeannie Albrecht. My thesis is entitled “Predictive Pod Auto-scaling in the Kubernetes Container Cluster Manager”, and it is entirely open-source. The written portion, presentation slides, and evaluation code can be found on my Github. Additionally, our contributions to Google’s open-source cluster container manager Kubernetes can be found on my fork.