Jump to: navigation, search

Governance/ElectionsSpring2012/Joe Heck

< Governance‎ | ElectionsSpring2012
Revision as of 19:03, 23 February 2012 by JosephHeck (talk)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Joe Heck

(also at JosephHeck on this wiki)

I'm candidate for election to the Project Policy Board and as Project Technical Lead to Keystone

OpenStack bio

I've been active in OpenStack since 2010, starting out working around helping with continuous integration, documentation, and some background efforts in the developing dashboard. I've continued to help with the CI effort, specifically around basic builds for Horizon and Quantum. I've added significantly to documentation both with the formal PDFs, the build process, and just updating and including documentation on this wiki and in informal venues (understanding FlatNetworking, updating Keystone documentation, etc). In the past 6 months, I've been focusing on collaborating with downstream packagers to update Horizon and Keystone, doing a great deal of work to drive Keystone into a stable and sustainable product, and helping prototyping out the initial vision for what Anne and the doc team just released as http://api.openstack.org/

I started out working with OpenStack as a consumer while Director of Cloud Services at Disney; more recently have been working on the side of providing OpenStack to companies working for Nebula.

Questions

1a. Since the last elections, what areas have you focused on and what contributions have you made in order to improve OpenStack as a whole?

Most of my focus since the last election has been on documentation and Keystone. I prototyped and helped drive some of the direction to generate a consolidated OpenStack API page with the Doc team that ultimately became and was recently released as http://api.openstack.org/. I also added significant documentation to OpenStack, and contributed to helping manage Keystone and Horizon as a core contributor on both - triaging bugs, coordinating blueprints, and code review and patches for both of those projects. I also continue to help those teams with CI knowledge - generally assisting the CI team with either of these projects and providing some back-up support there.

1b. Since the last elections, what areas have you focused on and what contributions have you made in order to improve your project?

Along with a huge amount of effort by Termie (Andy_Smith), and a large number of other folks, introduced and baselined Keystone to a new codebase. The effort, along with documentation changes and integration efforts, was driven entirely from functional and integration testing with an eye towards dramatically simplifying the codebase and getting the project to where it could be made sustainable moving forward in terms of both object structures and APIs, with the clear idea that Keystone is not an identity management system, but will used most frequently as a proxy to an existing identity system. Keystone needs to be a viable product to be run to support OpenStack, as all the other components should be able to be deployed to expect it's sustainable and continued use, at the same time the API needs to be carefully managed to not introduce unnecessary complexity in order to allow for it to function as a proxy to existing identity and authorization systems.

To be very clear, Termie did most of the heavy lifting, and regardless of the PTL elections I will continue to expect to look to him for guidance related to the code structure and continued managability.

2a. What are the most pressing/important issues facing OpenStack as a whole?

The nascent structure of the Foundation is the most important strategic issue, while stability and continued integrated quality improvements across OpenStack projects is the most pressing tactical issue.

2b. What are the most pressing/important issues facing your project?

Driving to clear, simple relationships for the code structure, API, and exposed resources (through the API) that Keystone provides as a proxy to allow for implementing role based access (authorization) and simple identity verification (authentication). Having a project that can be run as a viable product in a production installation of OpenStack is critical, as I expect many implementations of OpenStack will expect to use Keystone - sometimes by itself, sometimes as a configurable shim as additional backends are added. Other organizations will replace it entirely, hence the need for clarity on the API to match the implementation.

3. What is your relationship to OpenStack & why is its success important to you and/or your company?

OpenStack is important to both me and the company for which I currently work, but I think the more interesting question is why am I involved with this project. I don't work on OpenStack with just hours paid for by the company I currently work for (Nebula), although they graciously allow me to contribute a large number of hours during the day to the overall effort. I started using OpenStack while trying to drive down the cost of providing computing services for Internet sites across the Disney corporation. Even at a medium scale, many enterprises *can* find significant cost savings in using compute as a completely ephemeral resource if they invest and drive the systems with automation.

My career has jumped back and forth between development and operations - and driving more automation has been a consistent theme. OpenStack is the first (to me) truly open infrastructure as a service that provides a viable platform to push the commoditization of "making and using a virtual machine" to a new level. The value in OpenStack to me is making that layer stable, consistent, open infrastructure as a service, which in turn enables a whole new set of venues for making computing resources even more easily accessible and cheaper to run than before.

From that, I think it's critically important that OpenStack not just be an API specification, but a solid implementation (product) to provide a concrete interoperability story to drive this commoditization. Diversity in opinion and focus is critical to success, especially with infrastructure as a service, which brings together systems that have historically been very separate across almost every organization.

Finally, I've always been a fan of the concept that it is individuals in an organization that make a difference in this world, and open source projects are, to me, an expression of that concept.

Links