openstack team mailing list archive

[Ed Leafe <ed@xxxxxxxxx>]

	As some of you know, I agreed to propose, prepare and give a talk at next month's US PyCon in Atlanta. I felt that it would be a missed opportunity to have one of the biggest and most significant open source project in Python not represented at the largest Python conference. The talk was accepted, and it's entitled "Dealing with Concurrency in Large-Scale Systems". (http://us.pycon.org/2011/schedule/presentations/186/)

	I've spoken at PyCon and other conferences before, but always on topics that I knew inside and out. When I proposed the talk, I still hadn't begun working on OpenStack code yet, and relied on Soren and Eric to get to a point where I could write a cogent proposal. So I'm at a bit of a disadvantage when it comes to deciding what to focus on.

	That's why I'm writing this: I need those of you who know the code base best to identify the stuff that you've worked on that would be the most interesting to include in the talk. Remember, the audience will be a bunch of experienced Python geeks who will eat up cool or unusual approaches to solving technical issues.

	So what is "interesting"? I see several categories:

1) Scaling issues: what sort of design decisions were made to accommodate the large-scale demands of OpenStack? And what sort of atypical implementations were created to make these designs possible?

2) New techniques: stuff that if you had heard someone talk about at, say, last year's PyCon, you would have been impressed by. Sure, you're familiar with it now, after having worked on the OpenStack code base for a while, but try to remember the effort of creating that solution (if you were the one who did so), or the reaction when you first understood someone else's awesome work.

3) The unexpected. What did you run into that did not appear as an obvious choke point that ended up requiring a bit of work to overcome? These types of battle stories always resonate with fellow devs.

	Of course, I will give full credit for any suggestions I use in my talk. My goal is to get people who don't know much (or anything) about OpenStack to come away from the talk impressed by the incredible work being done to make this project happen. I can't do this by myself.



-- Ed Leafe