By Val Bercovici, Cloud Czar, Office of the CTO, NetApp
I’m just going to come right out and say it. Amazon Web Services (AWS) is the best place to start developing and testing any cloud application today. I’ll go further. If you’re the technology decision-maker in charge of a business transformation project sponsored by marketing, sales, engineering or customer support, it’s tantamount to professional malpractice to start anywhere other than AWS.
Amazon provides incredible value in terms of rapid innovation, quick iteration and fast-fail experimentation. In short, the economics of elasticity heavily favor Amazon-and increasingly Microsoft Azure, Google Compute Engine (GCE) and IBM SoftLayer. You’d be putting yourself at a competitive disadvantage by going anywhere else. That’s it. Speech over. Soapbox stowed.
Having said that, just because you’re born somewhere doesn’t mean you have to spend the rest of your life there. This is also true with cloud application lifecycles. In fact, the oft-heard claim that you can “do everything on Amazon”-or on Azure, GCE or SoftLayer-is naive. The reality is there’s no single landing point for born-in-the-cloud or cloud-first type apps. Most of them cannot afford to remain in the cloud forever. It doesn’t make business, economic or operational sense for them to do so.
While we’re busting myths, I’d like to outline a concept I think anyone making the move into cloud today-and DevOps in particular-will find highly informative. It’s a concept that shatters a lot of myths about what cloud is and isn’t. It involves using NetApp’s Data Fabric to follow the application lifecycle. I call it the “Lean Cloud.” It’s something I’ll cover extensively at the upcoming NetApp Insight user conference in Las Vegas.
Application Lifecycle Milestones
There are three major milestones in the lifecycle of most modern applications. Interestingly, these categories and percentages map to three tech spend ratios reported by CXOs and heads of IT, according to research by Gartner. But that’s for another blog. The three application milestones and their corresponding spend ratios are:
- Transform (14 percent)-often spent on app development and testing labs.
- Grow (20 percent)-often spent on production deployments and scaling of new and existing applications.
- Run (66 percent)-all about maintenance, usually of legacy infrastructure.
How does this relate to the Lean Cloud? First of all, let’s be clear-to optimize both your innovation and operational efficiency in the cloud, you have to go hybrid. In a DevOps context, this often means planning to develop and deploy your app across multiple clouds simultaneously. The Lean Cloud concept rests on the idea that you also need to optimally locate your app over its natural life cycle.
Economics of the Hybrid Cloud Ecosystem
I’ve already stated that the economics of elasticity overwhelmingly favor public clouds (AWS, Azure, GCE, etc.) for initial functional development and testing of apps. Elasticity, in cloud speak, is the degree to which a system or cloud layer autonomously scales its capacity to a workload over time. It has enormous value when workloads are very volatile and dynamic, as they are during the ongoing development and testing cycle. Elasticity can also be useful for initial deployments and sizing infrastructure requirements. This allows Line of Business (LOB) to implement a fast-fail approach to innovation with minimal risk and maximum agility. Most of these experiments fail, but a small portion-typically about 10 percent-become successful new apps, or new/updated modules (microservices) within existing apps.
However, something changes when one of these 10 percent of apps becomes popular and successful, when you try to scale it. Once an app is in production and its workload becomes predictable, the value of elasticity diminishes in favor of service quality and cost containment. You know what the workload is going to do; you know when there are peak loads and when there are not. At this point, a lot of customers move their successful experiments off Amazon (or other hyperscale clouds) to what I call Tier 2 service providers or managed service providers (MSPs), such as Verizon Cloud, T-Systems, Telstra, HP Public Cloud and Rackspace.
Tier 2 providers are not less valuable than the likes of Amazon. They simply offer a different value. They’re not as elastic, as dynamic or as flexible. They don’t have the same richness of development service. But they have business models, infrastructures and service levels that are more attuned to fixed or contained costs for predictable workloads. So for a given set of workloads, they can offer lower costs, higher performance, higher availability, greater security, greater regulatory compliance-whatever the business needs.
For mature or legacy cloud apps that have static or highly predictable workloads, the Lean Cloud concept comprises a third landing point. In order to realize maximum cost efficiencies, these workloads can be further migrated to hosting sites such as GoDaddy or DreamHost, or back on premise in the customer’s own data center. No service provider needed-just a decent IT staff. And, in the event of an unpredicted spike in app workload, you retain the option of bursting data on-demand back to a hyperscale cloud.
NetApp’s Data Fabric is the only hybrid cloud data management architecture that enables hyperscale (born-in-the-) cloud apps to simply and efficiently migrate data as their economics shift throughout their respective lifecycles. It provides the necessary choice, visibility, control and data mobility to take advantage of these shifting economics. Put another way, Data Fabric allows you to overcome the limitations of data gravity in the cloud. It allows DevOps to ride the Lean Cloud.
Have your registered for NetApp Insight yet? This technical conference provides customers and partners a chance to dive deep into the trends and technology shaping the future of storage and data management, starting with how to build a Data Fabric. Learn how to meet your data challenges from NetApp engineers and thousands of the brightest IT innovators. Register for Insight today!