April 15, 2011
This week’s announcement by VMware of its CloudFoundry PaaS product was an exceptionally refreshing surprise. Other have commented that it’s a rare thing for a big company to really do things right – but with CloudFoundry, VMware have ticked every box – opensource? Yes, avoids lock-in? yes, value added? yes.
A quick recap of what CloudFoundry is, and does;
- Open PaaS
- Supports apps in Java, Rails, Node.JS
- MongoDB, MySQL and Redis support
- Use it public, private and cross infrastructure
- No lock in (did I mention no lock in – this is BIG)
The comments among the clouderati were ebullient to say the least. Even the most cynical of souls, the ex-president of the private cloud himself, Christian Reilly, found himself getting excited;
The comments from the blogosphere have been very positive – Krish (who has been testing CloudFoundry for some time) loves it, hell even Gartner, not normally quick to jump on latest developments, is positive. I see this through two different lenses, so here they both are;
For developers, end users and IT staffers everywhere
This is awesome – PaaS IS the future of cloud, providing more value than IaaS and more flexibility than SaaS. PaaS that gives users the ability to migrate, that lets them use their stuff wherever they like is truly golden. This is a big moment
For the vendors
This kind of hurts. No one has ever made any money from OpenSource (deliberate provocative statement that is very over stated but intended to get people thinking). We’ve seen a real commoditization of infrastructure over the past couple of years. That kind of culminated a week or so ago with the OpenCompute project launched by Facebook which OpenSourced their data centre design – great for some, but also creating yet another commodity layer with little opportunity for differentiation.
There is every chance that CloudFoundry (and the offerings that other vendors introduce in response – because they will) will further commoditize the market place – only this time higher up the stack where there was still money to be made. Simon Wardley in his inimitable style suggests that this is in fact the intention of VMware – Wardley couches this in terms of a;
huge moat devoid of differential value in the platform space and a vast ecosystem driving this
If this is the intention – I worry about the unintended consequences of commoditizing something ABOVE the layer in which your bread and butter lies – VMware runs the risk of cutting off it’s nose in spite of its face (cue a dodgy metaphor).
So… where does the real opportunity lie. Well as Wardley points out, the real money comes from the high touch stuff – brokerage, marketplaces and, ultimately, service. It’s the very reason Rackspace has invested so much in OpenStack (disclosure – Rackspace is a client) – they see the futility of fighting against a juggernaut of commoditization and rather look to steal the edge by getting in early in the services differentiation game.
CloudFoundry, like OpenStack before it are simply beautiful. Open is good. Moving up the stack with open is even better. I love what it means for the industry and dearly hope that vendors can see ways to avoid fighting a commodity battle but instead determine paths to providing customers with real value built on top of these commodity platforms.
To the cloud (oh yeah, I think someone already took that line). Ah well.