I have been fortunate enough to be a week into a fortnights vacation away with my wife and kids at our holiday home in Spain. So right now as I look out across a beach to the sea with Gibraltar in the distance and the temperature dropping from its afternoon highs, nursing a cold beer, I've been able to fully catch up with every video released from OpenStack Summit in Paris which I wholly deliberately chose not to attend for once. I needed to recharge batteries badly and it was my genuine health or the insanity of Paris and the hubbub and noise of overfilled rooms and mass lunches and rain vs a hot sandy beach, long drinks, my amazing (decade younger) wife in a bikini and Spanish beer and food. The latter won. I make no apologies for wanting to take two weeks off for the first time in twenty five months. Anyone wanting to argue the difference needs to understand I can drown you in a the shallow end of any of the resort swimming pools as soon as look at you, after this week, I have practiced my technique. I am Aquaman of the Marriott set.
So over to OpenStack Paris 2014. The view from my sun lounger.
I've watched from afar like a demon this week. It's been great to catch up online and watch all the sessions. Actually when you're at the Summit there are sessions overlapping so you can't be in all places at once and while not all content was video'd and online the content that is there is superb. Hats off to the foundation for getting it there so fast too. You can watch the content here. Congratulations to the Red Hat team for getting so many talks accepted and the delivery of the content.
One thing is very clear, there is still enormous drive, passion and mass determination to make OpenStack releases qualified successes. Nobody can detract from the earnest efforts of all parties no matter who the contributing employer is.
First shot across the bows - my boats bigger than your boat...
Let me get one thing very straight from the get go. One thing I was very glad to see this Summit from the videos and decks that I've seen as a remote watcher. Paris seems, somewhat thankfully, to have had a lack of the marketing BS that has become so prevalent with the constant who is the highest ranking contributor to the project as a whole. We're professionals no matter what tshirt or cap or hat you wear and who pays your salary. We're chasing a common goal and waving willies about in public to say who is the biggest or who is the best is just incredibly poor taste and detracts from a lot of the interworking and common core goals that the OpenStack Foundation are attempting to deliver. This is about good code, influencing major adoptable change in how we help people get the infrastructure that fits their cloud business case and frictionless IT. If you are scoring points what else are you missing when it comes to understanding what first world enterprise IT want ?
There are distributions out there, I work for the company who are trying their hardest to make sure it delivers what the market expects building on years of enterprise experience with Linux and putting the best engineering talent behind that gains recognition from the markets wanting to trial and consume it. Those consuming enterprise customers markets don't need or appreciate a poorly conceived marketing slide that is at best oneupmanship, at it's worst just a 140 character land grab, it has no place in thought leadership. Period. Don't tweet me or send me a deck or marketing swag with it on or I'm getting the elephant gun and my steel toe capped boots on and going hunting.
So now I've got my pet moan out of the way lets talk shop as an external watcher perceiving how the world is consuming OpenStack latest greatest in bite sized chunks.
First up, keeping it simple
OpenStack has aggressive release cycles, has a multitude of sub projects and a host of goodwill and contributed code that deservedly allows it to rank as the leader of the upper echelons of Open Source goodness. A shining beacon of how to do things and achieve both success from a release and maintenance perspective but also of marshalling talent and consumptive code contributions from individuals, companies and projects to come up with a release cycle that is hard work to maintain against. My congratulations - and genuine admiration - of those involved many of whom I know and respect hugely can go on the record here as it has in presentations and podcasts I've released all over the world.
One issue is perception by the watching consuming public and the enterprise architects and that is the need, the fundamental core principle of keeping it simple. Plainly put Icehouse and Juno are still seen as rocket science to many in core consumable released non supported format. I watched one video from the team at Rackspace that called it exactly right and I hope that it gets some airtime and credit as it was right on the money as a call to arms for the Foundation and the maintainers to get to a point where ease of use has to be a mission statement. Fostering ecosystems is critical, bringing functionality into the core is a constant need.
Do we need any more plaudits ?
Not knocking Jim Zemlin in his keynote as anyone else blowing hype and sunshine up our combined asses as to how big "a blockbuster" is simply needs to stay home, the last thing you need to do in a room full of excited OpenStack types is to pat mutual backs and inflate already inflated egos. What's actually needed is more critical leadership around concentration on maybe looking to increase the width of the release window (six months is overly aggressive and actually makes new adopters shrink back in fear) and to educate and mentor maintainers of sub projects as to the needs to increase the fundamental ease of use of their functions and core capabilities.
Also - are we solving the problems that actually are relevant in the marketplace ? Are we moving at such a pace that we're not engaging with required functionality and getting instead the sexy stuff like SDN in there because it's en vogue ?
Right now, today, I see more, bigger, mainstay companies who have deployed Apache CloudStack over OpenStack and these aren't small organisations they're big companies, because it does what they need to and it doesn't terrify the life out of them. If OpenStack is going to be in that same vein, remembering that the Apache CloudStack ecosystem combined with the clueless parent company has 0.1% of the mindshare and the groundswell push behind it then it has to do basic stuff better. Some of that is packaging, fairy dust, documentation and means that Foundation and contributors need to engage better with their future consumptive masses.
For us, we take what is out there, polish it, build it into a supported product with a pedigree and core function behind it and deliver it to people who want to feel safe. If I was a CIO today looking at OpenStack I'd want to match core fears of "being able to keep up" with a comfort factor of having something supported. At the core the functionality and the best practices need to tighten up to allow OpenStack the success it deserves. Theres a very real chance it will miss a high percentage of its goals if it doesn't listen.
My genuine admiration and worthy applause goes out to all speakers, panellists, those on booth duty and who took time out to attend. Me, I was on a beach with a beer. Genuinely, right now as yet another cold beverage disappears and the light fades over the white stucco plaster of the houses here in Estepona on the Costa Del Sol I need to be convinced that catching up via YouTube, Twitter and the polished editorial of Steven J Vaughn Nicholls et al isn't a better way to do Summit than fighting for a seat in a crowded room and queuing for a mass meal with 4600 other attendees.
Kudos to all of you who did make the effort to go.