Puff on a hybrid – next thing you know, you're hooked on a public cloud
Cost barriers are falling fast. Go public, it's inevitable
Blocks and Files I had a flash of inspiration today. Hybrid public-private cloud systems are becoming a gateway drug to pure public clouds. Why is this an arguable view? Let’s look to the ideas discussed within The Big Switch, written by Nicholas Carr.
In his seminal book, Carr argues that a public IT utility provider will provide CPU cycles and storage of terabytes at a lower cost than someone kitting out their own private data centre because the utility provider benefits from a greater economy of scale.
A good analogy to this is comparing a corporation operating power stations to businesses having their own generator in the basement.
Hooking public cloud services to a private data centre makes storage and workload migration to the public cloud easier. If you can move workloads reliably and simply between private and public clouds then the risk of things going wrong when moving them to the public cloud is reduced.
The public cloud world consists of more than Amazon Web Services, Microsoft Azure and Google. Most telcos are implementing and selling some form of cloud IT services, and many are enterprise class. When they offer workload importation and workload operation with SLAs and protection then their price tags can be compared to in-house IT costs.
If you don't like the gateway drug analogy, try this: building hybrid clouds is like breaching the enterprise IT dam and letting its contents flow away.
Although people like NetApp’s CTO Jay Kidd can argue that Amazon’s storage services aren't cheaper than NetApp’s arrays over a three-year period, I think that is too simplistic a view.
What is the cost of acquiring, implementing and operating a NetApp array when you factor in everything, from the power and cooling to the management of the admin staff? After all, you don’t buy and operate a storage array in isolation; it runs inside an infrastructure that has to be paid for and some portion of that cost should be allocated to the array.
That cost barrier to the public cloud is, I fear, going to melt away. In fact it will probably reverse, with enterprise IT becoming visibly more expensive than the public cloud, especially amid the ongoing price war in the public cloud marketplace.
Breached: Will your data pour away like the water from the Möhne Dam, breached by the famous Dambusters raid?
In my view, an enterprise’s IT estate can be likened to a fried egg. The yolk is the set of data centres and the workloads they run. The white is the entire set of endpoints in the enterprise.
As an enterprise’s IT estate is gradually lifted into the public cloud, you're left with a ring of end-points with no centre. Just end-points talking to public cloud resources.
This will take a decade to come to pass. Possibly longer, but there is beginning to be a feeling of inevitability about this.
If enterprise IT suppliers provide hybrid cloud resources, they are trading short-term gain for long-term loss. Amazon, Google and, soon enough, Verizon and its ilk, will not buy costly branded arrays or servers from EMC, HP, IBM, HDS, etc. They’ll follow Facebook and buy in white-box stuff from factories in Taiwan and similar geographies.
Present enterprise IT kit makers are building a one-way street leading to the death of their hardware-based businesses by supporting public cloud access from private clouds. The hybrid cloud is a trap, a trap that will close on the very suppliers who are building it.
Amazon, Google, Azure, Rackspace and the telco public cloud service providers are not their friends. Strategically, they are their deadly enemies and the mainstream IT product suppliers of today had better figure out how they will supply hardware, hardware-plus-software, or software-only, to the public cloud service providers in the future; a future that is coming towards us and them the faster as more hybrid clouds are built. ®