Joshua Harlow on Task Distribution
Get an expert opinion on the concepts you learned in this chapter.
We'll cover the following...
Asking an expert
Following is a conversation with an experienced Python developer Joshua Harlow on the topic of task distribution.
Joshua Harlow
Q: Hi Josh! Could you introduce yourself and explain how you came to Python? |
---|
A: Well hi there! I grew up in upstate New York. I went to school at RIT (and prior to that Clarkson University as well as a NY state college) and graduated in 2007 with a Masters in Computer Science. During this time, I interned at IBM where I did some automation work using Jython and Intel, where I helped the graphics team by interconnecting Ruby and C#. While I was in college, I got very interested in distributed systems, and the interconnect/potential when combined with AI as well as a stint in language theory and applications. |
After graduation, I came to work at Yahoo. After working on various projects such as the homepage (www.yahoo.com), I got recruited into a sub-team under the CTO organization where we were tasked with determining the cloud solution Yahoo should invest in and use. OpenStack was a nascent open source cloud technology back then, but it was what we thought had the most potential. Since OpenStack was being written entirely in Python, this is where I got my actual initiation into Python. Over time, I have come to enjoy Python, come to learn it deeply, been featured in a book on it and never looked back! |
Q: What’s your experience with building large scale systems? |
---|
A: I have been working in the industry for around ten years. Two-thirds of that was working at Yahoo where I was introduced to various patterns and systems around how to design a scalable system. At Yahoo, anything that runs must be able to scale. Part of my time there was being involved in the OpenStack community while also being one of the key Yahoo OpenStack contributors and one of the technical leads/architects on the larger team there. I have tried to work, inject, or at least share some of those same ideas/lessons learned into various OpenStack projects. |
Before this, I did graduate work using a highly distributed paradigm called agent-oriented |