This post was overdue by now - but today's (July 24th 2013) events triggered it into existence, as I wanted to write a post on the OpenStack dynamics as we see them unfolding in the coming quarters since a long time. Cloudscaling's Randy Bias' unbiased (pun intended) open letter to the OpenStack community was just the spark that was needed.
OpenStack turns 3
With all these babies turning toddlers - we have plenty of 2 and 3 year olds around now. Time to take stock. And specifically for OpenStack - the toddler had a growth spurt of massive proportions in his 3rd year with IBM and HP joining the party.
For a real world toddler that would be like being given an Olympic size swimming pool and half a Disneyland as personal play area. Maybe exaggerated - but if you consider the resources and scale that IBM and HP brought the OpenStack 2 year old - the analogy may not seem so grandiose after all. And I am not mentioning what the joining of RedHat and VMWare would mean for the 2 something year old.
What started as little less than 30k lines of code with the Nova compute project (from parent NASA) and the Swift storage platform (from parent Rackspace) has now ballooned into a 600k lines of code 3 year old, that also dabbles in dashboarding (Horizon), image storage (Glance), identity (Keystone), networking (Quantum), block storage (Cinder) and many more activities.
Cloudscaling CTO and co-founder Randy Bias stirred things up with his open letter to the OpenStack community. There he states that practically AWS has won the battle for innovation in the cloud - out executing the OpenStack conglomerate and that it's time to honor this reality. Basically OpenStack should focus on integrating the on premise data centers with public cloud offerings, bowing to the Amazon APIs. Bias adds how APIs got somewhat hijacked by Rackspace in the early OpenStack life years - so the OpenStack community should not treat APIs too much as holy cows.
And while Bias is certainly right on the Rackspace influence, we think he underestimates the unwillingness of the big boys in the OpenStack community to let AWS get away as the winner. IBM, HP, RedHat etc all did not join OpenStack to declare AWS as the public cloud standard (and winner) but to form a broad alliance against the GAMO (Google, AWS, Microsoft and Oracle) vendors. And in their mind - they haven't even started to compete - it takes enterprises like IBM and HP months to get resources shifted, allocated and productive. For them 12-24 months are necessary to even see what speed the 3 year old can develop - being put on an enriched diet, more exercise, more teacher, enrichment etc (ok, the analogy gets quirky here).
The OpenStack Core Dilemma
The key takeaway from Bias letter is that the core problem of OpenStack are the close to their heart interests of each of their members. While Bias blames Rackspace correctly, we should not let him get away, that an embrace of AWS (and GCE, which he mentions, too) - would certainly help Cloudscaling to monetize some strategic investments. And the same is valid for IBM, HP, RedHat, VMWare, AT&T, Cisco etc - all of them have their own interest, strategy and path for differentiation - and as they hope - domination of the cloud going forward.
The interesting question for the observer is - how long will the OpenStack community play nice with each other - and when will the gloves come of? This is very hard to predict at this point, but we think that the easy life for the 3 year old will continue for 1-2 years - as the rich uncles and aunts (IBM, HP etc) - just get serious to invest in the 3 year old. They will play nice for some time and then may get into a fight for the elementary school the 5 year old should join for kindergarten.
And likewise - should any major player in the OpenStack community see the chance to step up to the GAMO players - we do not think they will hesitate for a second. But they would have to make sure they can pull it off and break free from the rest of the pack - as they risk to loose cherished inter-operability that they have been promising and selling to their customers.
This reminds me of the situation of an escape group in cycling (yes maybe too much Tour de France recently) - this group needs to work together to keep the peleton at bay (here the GAMO vendors) - but if any of the riders in the group thinks he can win - he will go for it and all cooperation is so passe.
It all plays into AWS hands
With all uncertainty that can be created in the OpenStack community for now and the near future - it all plays in AWS hands - and to a certain point in the rest of the GAMO players - as they offer a clear path and roadmap to the future. And it's this reliability and track record that enterprises ultimately make the foundation for their cloud investments.
And while e .g. IBM and HP have significant trust credit in the market and with their customers, they still need to earn their trust specific to the cloud - something they are lacking in comparison to the GAMO players. So it will be interesting to see how the adults in the Openstack community will be able to keep the hot headed members in check - for their mutual benefit.
The noise levels in the OpenStack community are up... and the potential cracks form this noise will only make the core argument of Bias' letter - the AWS leadership - a stronger one. But the desire of the large contributors to OpenStack to create a counter balance to AWS is and will remain very high for the foreseeable future.
It will be interesting to see if those enterprises (the IBM, HP, Cisco, AT&T etc) will be able to gloss over these noise levels and get the OpenStack community to not only deliver but also become so competitive towards AWS and the other GAMOs, that they can take market share from them. Exciting times ahead.
P.S. And for those who are wondering - quo vadis - latin for where are you heading / going?