Is there a way to succeed with OpenStack?

Several experts and non-experts such as me have criticized lack of focus and leadership in OpenStack efforts. It often felt like they were playing the wrong game, chasing wrong enemies and were running around in circles. To add to this, several folks have highlighted problems with the technology itself. Technical problems can and will be fixed, but headed down wrong road because you had a bad strategy will sting quite a bit.

Is OpenStack done? Not really, I do see a path forward for (large) companies to succeed and make money with OpenStack. I will outline my thinking here.

First few observations:

  1. The game for the future is Elastic Cloud – provide (perceptively) infinite compute, storage resources as a start for your customers and do not let them experience limits of infrastructure that traditional approaches surface.
  2. Private Cloud is a flawed thinking that will run its course soon.
  3. This requires willingness to offer low cost offerings and compete on price for commodity infrastructure resources.
  4. Realize that it’s actually not a zero sum game. You don’t only win if rest of the market loses – it’s still an early enough market that you should learn to co-exist.
  5. Accept what AWS has done is disruption to your traditional model – they are not competing on technology, they are competing on disrupting traditional enterprise software business model
  6. Responding to disruptive innovation is taking a harder look at your Resources Processes and Values (refer to Clay Christensen’s RPV theory).
  7. The market is ready to move beyond integrated offerings and may be getting hungry for open architectures when it comes to IAAS.

If you were to accept my observations, here is what I would recommend as a course of action ( using OpenStack as technology underpinning ).

  1. Create a separate BU with a GM to focus exclusively on public cloud that will be competitive with AWS and in the process will cannibalize your own private cloud offerings.  Give this BU very limited budget. Why should you have a separate BU? Refer to RPV theory again. Take it from me – I have failed in my career to do it within existing structure and most recently a half-ass BU structure.

One note on cannibalizing: Don’t worry about upsetting existing holy cows in your organization – they will be disrupted – only question is whether it’s your own BU or competitors – if your BU does it, these holy cows have a chance of retaining their jobs and retire with dignity.

This BU shall:

  1. Go with a pricing structure that is on-par with AWS and GCE. Sell this offering independently; create a separate, small and dedicated sales team. They should be commissioned on selling this offering only, nothing else.  (A major chunk should come from on line sales though).
  2. Make interoperability possible from your public cloud to AWS, GCE and Azure. Highlight that your offering is open, while rest are closed. See #7 above.
  3. Create the best interoperability possible with OpenStack deployments in a data center environment, but encourage those customers to get on a plan to move to the public cloud.
  4. Refuse to accept constraints such as artificial certifications, release processes that old timers at your mother ship may want to impose on you.

I do not believe you are at a disadvantage because you are starting late – this may be right time to offer a modular open option for customers growing weary of integrated IAAS offerings.

Advertisements

7 thoughts on “Is there a way to succeed with OpenStack?

  1. Boss,

    While in theory this is good, solid advice, there is one fundamental flaw with your thinking: Namely that the very strategy you suggest was tried already by really smart men more than 2 years ago – and failed.

    As unfortunately I’m inexcusably sober right now and this story still pains me , let me head to the closest bottle of Ardbeg before typing an answer ..

    To be continued…

  2. With the clarity of vision of one bottle of Ardbeg (and counting…) and corresponding factual inaccuracies (for fun and literary purposes, as it were….) here comes “OpenStack as a technology behind public IaaS — Drunken History #37”:

    Once there was a golden fairy that became the ruler of The Light Blue Kingdom. And when the fairy came to rule, she looked up to the skies. And the fairy saw the sky was scorched by big bad AWS burning everything in its way, with worse yet to come. And then the fairy looked down, and saw mighty dinosaurs ruling the fiefdoms with a mighty roar, totally ignorant to what was coming their way.

    So then the golden fairy took an enchanted unicorn with a turban and whispered in his ear: “You will be my enchanted saviour and source of our future competitive advantage. Our kingdom is vast, and great resources are still to be found for the right seeker. So here is a nice budget and the keys to our kingdom– go invent us a future, away from prying eyes”.

    So with the blessing of the golden fairy, the enchanted unicorn proceeded with great haste to do achieve that noble goal. He chooses OpenStack — a promising young new technology — to deliver “AWS-like”, public IaaS. Then to do that proceeded to build a great team consisting of the best and brightest money could buy. And all worked tirelessly do achieve that.

    But all was not well. The technology was not even premature — it was “pre-alpha alpha”. So our unsung heros, best and brightest as they were, had to move mountains to deliver a working service — but with delays after delays after more unforseen delays…. And to make matters worse, there was the plague of perfectionism — ’cause how would someone justify being hired as a rockstar other than by proving to others that he is indeed that by taking extra-time to create super-awesome technology ?

    And there were more problems: As our hero unicorn fought hard to build this future from nothing, and rockstar perfectionism, and technology shortcomings, and…and… and….there was much gnashing of teeth by the dinosaurs, much unhappy by what the unicorn was doing. So they were stomping their ground and spewing much FUD , with their mighty multi-billion dollar growl of “our established business is under internal threat — unacceptable !”. And there were insidious (as they were real !) objections that actual results are much below the original hype — just as our prophet Geoffrey A Moore told us (“you always initially overestimate that size of the market”…)

    And then there was fate: The golden fairy was no Lou Gerstner — as precious few ever are. And neither did she read (or was ever told of…) Geoffrey A. Moore “Escape Velocity” to realize that was _her_ responsibility to drive this tirelessly with a steamroller across the realm. And that there’s _one_and_only_one_ disruption one can make “mainstream” at a time, and that can be done by one person and one person only — herself. So when the unicorn was “delegated” to do just that (with the best intentions..) it was the kiss of death: Not only he was distracted from build that bright future that he was first asked to do, he had to preach & sell it to the dinosaurs too… So when he inevitably fails in that impossible task, our unicorn shakes his shoulders saying “Responsibility without control is a suckers’ game”, restricts back the focus to what he was supposed to deliver to start with, and leaves the rest of the kingdom to burn as it would inexorably do anyway….

    And then there was OpenStack: Not only the technical problems were big, technical egos and pissing contests in the ecosystem even bigger, but the landscape — and narrative — changed as well. And then there was a Foundation. And lawyers. And polictics. And cries. And facepalms. And laughter. And then even more cries…

    So, eventually the golden fairy lost her patience, and her spirit broke. And blamed the unicorn for failing to deliver to the original hype. And the unicorn was much saddened, and threw in the towel with a very heavy heart…

    And then the rest is tragic comedy.

    So boss, please do me a favor: Whenever a thought leader tells you “OpenStack failed to be the AWS clone it should have been” tell them two things 1) Yes 2) Fuck off.

    Now, what’s the future instead ? Well, now that’s for another bottle of Ardbeg to reveal….

    /Azamat

  3. Respect bro – as always. I have to tell you my side of the story over drinks – exact scenario – wasn’t a new fairy, it was same fairy whose heart grew weak.

    I was reminded of this quote reading your experience:

    “It is not the critic who counts; not the man who points out how the strong man stumbles, or where the doer of deeds could have done them better. The credit belongs to the man who is actually in the arena, whose face is marred by dust and sweat and blood; who strives valiantly; who errs, who comes short again and again, because there is no effort without error and shortcoming; but who does actually strive to do the deeds; who knows great enthusiasms, the great devotions; who spends himself in a worthy cause; who at the best knows in the end the triumph of high achievement, and who at the worst, if he fails, at least fails while daring greatly, so that his place shall never be with those cold and timid souls who neither know victory nor defeat. ”

    Now, coming back to business, OpenStack shall never be AWS clone – it has to be better. There are two dimensions where it can be better:

    1. Open architecture
    2. Interoperability with other Cloud offerings

    #1 is very possible. #2 would require walking over lot of eggshells in a fort called Foundation. Those eggshells are fragile and breaking egos is fraught with lot of Danger.

    There are handful of companies that can do this – couple that have sensible CEOs, but not sure if they have the courage or grit required to pull it off.

    • Much truth in that bro – respec’ (as always…)

      To your two points – esp the 2nd:

      That rabbit hole goes deeper – much deeper. Done properly (“properly” == As visionary as enabled by one case of Ardbeg) it would mean redefining what IaaS means & how it works. Ergo break of much glass with the existing OpenStack merry-fuckyou-tocracy. But here’s the catch: IF it succeeds, it renders the whole debate (2nd point) moot, as you just redefined a whole category via “grass-roots gone viral” (aka “_if_ you do OSS and want to be the odd one out, WTF don’t you do it properly” ?)

      As a case in point, at our the next Ardbeg meeting I will bring one bottle of Aberlour A’bunadh as an example & discussion starter.

      /Azamat

    • Please see #1 – the game is one of elastic and (perceptively) infinite access to compute and storage resources – this is not possible with private cloud without committing large upfront CAPEX.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s