An Open Letter: The OpenStack Foundation Should Be a Level Playing Field

Joshua McKenty  /  Mar 9, 2012  / 

I’ll be the first to admit that I’m not a diplomatic person. When we were launching OpenStack, this was a bit of an advantage (if we had waited for permission before releasing the Nova source code, we’d still be waiting) – but since the first summit, the community has grown so quickly, and become so diverse, that I have tried to leave discussions of governance, foundation structure, dispute resolution, and most particularly monetary corporate contributions, to others with more… tact.

But now I feel I have no choice but to speak up; I’m deeply concerned.

The biggest, splashiest openstack stories of the past two years have all had the names of huge, multi-national corporations in them – names like IBM, AT&T, Dell, HP, and CISCO. And while their participation has been tremendously positive for the project (with Quantum and Crowbar standing as examples of this), I see things trending in a direction that makes me nervous for the smaller players – for the startups who will live or die on the strength of the OpenStack project. Like Piston Cloud.

The current official proposal for the foundation creates a new class of super-members – with a sticker price of $2.5M (due up front) that puts it out of reach of all but a small handful of organizations.

This is not a new idea – it was the first structural proposal for the foundation that I heard from the organizing team, and I have argued against it (at times seemingly successfully) continuously since last fall.

I understand why it is appealing; it creates a small and manageable board of directors, with a large pool of resources, who shouldn’t have too much trouble guiding and directing the outcomes of OpenStack. But it’s not a structure that represents or embodies the principles that OpenStack was founded upon, and I think that while it may offer some short-term benefits, it may be damaging to the long-term health of the project because it strangles the ecosystem of contributing companies we’ve worked so hard to create.

The “right” structure is a much harder thing to organize:

  • It recognizes and requires project contribution (code, tests, docs, bugs and evangelism) along with cash
  • It has a single class of corporate member, a level playing field
  • It has room for non-corporate members in the meaningful governance bodies (not tucked away in ‘advisory’ boards)
  • It aggressively and publicly resolves the conflict-of-interest between the ‘company hat’ and the ‘project hat’

My understanding of the key challenges of this foundation board are the following:

  • Keep it small enough to be manageable (21 directors or less)
  • Supply enough funding to carry on with most of the current project support activities
  • Ensure representation of the diversity of the OpenStack community
  • Provide a mechanism for “industry luminaries” as well as OpenStack users and consumers to provide input and feedback

The target budget of the Foundation is around $3M per year. Without getting into a discussion about whether that’s reasonable or not, I’d like to brainstorm how we could reach that goal in a way that better reflects our goals for an open and democratic community. How’s this for a proposal:

  • One class of corporate member
  • Provide reasonable evidence of 2 FTE (full time equivalents) working on OpenStack in some capacity
  • Commit to 2 years of sponsorship, on an evergreen basis, but paid annually
  • Individual members, if there are any, cannot be employed by a corporate member

My rough calculation, having a reasonably good grasp of the interests and level of engagement of the various corporations in the OpenStack ecosystem, is that we could expect around 15 of the 150 companies involved to meet these requirements. $3M divided by 15 = $200,000.

It’s a high playing field, but at least it’s a level one. It doesn’t change the structure or composition of the technical committee, and it doesn’t limit the ability of the foundation to raise money in other ways (sell sponsorships for events, charge admission for conferences, even license the use of the trademark for training or certification).

If we have a simple pay-to-play model, then we can trust market economics and enforce transparency of spending. If we have a simple “meritocracy”, then we can expect the most skilled and dedicated to rise to the top, provided we’re extremely careful about how we measure skill and dedication. If we blend the two, I’m deeply concerned that we’ll see the worst of both systems play out over time – the selfishness of market-driven economics dominating our decisions with the petulant moralism of the meritocracy. Hoping for any other outcome is, in my opinion, foolish optimism.

At the core of OpenStack is the idea that a single project could address the needs of ALL of our organizations – large, small, producers, consumers, non-profits and tool makers. We need to guard that vision, and protect it from our best intentions. No one in the community, whether individual contributor or corporate sponsor, can claim to speak for (or even understand the perspective of) the majority of us. We’re simply too numerous, and too diverse. If you believe, as I do, that *your* company should have a stake in OpenStack’s future, then now is the time to speak up in favor of the level playing field we originally set out to create.

With (attempted) diplomacy,

Joshua McKenty
CEO, Piston Cloud Computing
Co-Founder, OpenStack

Share this:

Posted in: Blog


  • http://cloudscaling.com/blog randybias

    Cogent as usual, Josh. Cloudscaling thinks this needs more real discussion. http://www.cloudscaling.com/blog/cloud-computing/open-communities-deserve-open-communication/

  • http://www.collierclan.net/mark/ Mark Collier

    Hey Josh – I responded to your proposal on the mailing list as well (I think it’s identical to your blog post)? but I’ll add it here as well.

    Thanks for taking the time to consider the proposed structure and provide your thoughts on it. I don’t think your concerns are very far off from what we are wanting to address with the Associate Member category, which allows more companies to be represented at a lower price point. The current funding proposal is on the wiki for reference: http://wiki.openstack.org/Governance/Foundation/Funding

    Let me address a few of your points regarding the proposed plan:

    Accessibility: Associate Member fees start at $50k/year which is substantially lower than a $200k/year flat fee model. The members will elect resprenstatives as a class to the Board. Those Board members will have the same roles and responsibilities as any other Board member. Additionally, companies who wish to show their support for OpenStack may also become a sponsor without paying the full Associate Member fee. Sponsor levels have not been worked out, but I suspect they will start in the 20k/year range.

    Board Size: My sense is that with a flat fee of 200k/year, we would still get many more than 15-20 interested parties, which would quickly overwhelm the board, making it unmanageable. The Associate Member concept mitigates this issue by allowing the class to grow (for accessibility), without the seats balooning, by having elected representatives. It also reduces the risk of shaking up the board makeup and size through acquisition.

    Strategic Member fees: One small clarification: the proposal calls for Strategic Members to make a commitment of $500k/year for 5 years, paid annually (not up front). This figure was driven primarily by the need to arrive at a reasonable board size, while also raising substantial funds for foundation operations.

    Strategic Member dominance: Board members who are appointed by Strategic Members will make up only 1/3 of the board, and will have the same roles and responsibilties as the other board members.

    Requirement for Strategic Members to have full time staff: The current structure proposal states that Strategic Members must “Provide dedicated resources (e.g. developers, legal resources)” and the funding proposal states that “…the general expectation is a resource commitment that is consistent with staffing two full time equivalents” so it looks like we are on the same page re: 2 FTE requirement.

  • Mike Milinkovich

    Josh,

    FWIW, if you want to see an example model that is known to work, take a look at how the Eclipse Foundation is structured[1]. Dues are tiered by company size (revenue), and there are resource commitments at the board-level of membership. There are elected members from both the lower-tier corporate members and the committers. The number of elected seats grows as the number of strategic members goes up to ensure an overall balance. I can testify from personal experience that those elected seats are very important to the decision-making.

    Our board peaked at 28, but is now at 19. Obviously a big board takes some effort, but it is definitely manageable.

    [1] http://www.eclipse.org/membership/become_a_member/membershipTypes.php

  • Pingback: An Open Letter: The OpenStack Foundation Should Be a Level … | cloudguys.net()