Cloud computing is one of those innovations that opened a door through which has flooded even more innovation. Originally, infrastructure-as-a-service (IaaS) offerings built on server and storage virtualization enabled rapid deployment of virtual servers. As IaaS has matured, though, it’s become clear that infrastructure virtualization—via public or private clouds— simply steps one in a longer journey that brings even greater benefits.
For a couple of months, there have been blogs about hybrid clouds—enterprise computing environments that seamlessly link a private cloud running in your own data centres to one or more public clouds. Here is an explanation on how hybrid clouds provide a platform for application innovation that promises to bring a new generation of apps that can deploy quickly, scale without bounds, and achieve resiliency that would take enterprises months and millions of dollars to create using traditional IT infrastructure approaches.
While IaaS provides self-service for IT infrastructure teams who can instantly respond to requests for new servers, progress immediately slows to a crawl as app teams provision the new server with the software infrastructure required by the applications—database, Web server, application server, etc.—and then deploy the app itself. And because it is a different deployment platform, putting the app into production requires modifications to test and deployment processes and continuous integration and DevOps programs.
The corner that has been turned is that hybrid cloud-based infrastructure is bringing the benefits of cloud computing to application development teams by allowing them to develop apps without needing to provide for—or even know—the eventual deployment platform. Once developed, applications can be deployed into public clouds or into on-premises private clouds that share the same architecture, governance policies, development and deployment tools, and management and automation frameworks. And they can be redeployed as economics or business needs change.
That was the experience of Intel’s own IT organization when they implemented a private cloud within our data centres. They began by offering internal IaaS services and quickly fattened the platform by adding platform-as-a-service (PaaS) and database-as-a-service (DBaaS) to create a more standard app environment. They soon found, however, that they needed to completely change the way they thought about their computing environment. Rather than build the environment from the infrastructure up, they adopted an application–down approach that completely abstracted the infrastructure to enable an application stack that could deploy anywhere. The objective is every app running in the right place, where the right place is driven by the needs of the business.
Intel IT achieved infrastructure independence for its development teams using open source PaaS software designed to work across the entire app development lifecycle. It lets developers focus where they provide the most value—application functionality—rather than having to deal with the underlying infrastructure. To date, Intel IT has hosted more than 350 applications and 3,500 app instances on the new platform. By empowering developers, they improved time to market, reduced costs by 60 percent compared to IaaS services, and are operating at half the cost of public cloud services. They’re currently in the process of rationalizing a portfolio of 2,000 applications to determine which should remain in place, be re-hosted or redeveloped, replaced with SaaS solutions, or retired.
Application platforms built on cloud-native technology can bring benefits a traditional IT approach might take years to develop: Unlimited scalability, fault resilience, and dramatically reduced time to market. They let you achieve better architectural compliance for better governance and security and lower total cost of ownership.
The next wave of applications will be built on emerging cloud-native technology, but even now, solutions are available to let any enterprise take advantage of cloud-native application technology—often without even needing to redevelop the application. Solutions like Cloud Foundry*, Stratoscale*, VMware Cloud Foundation*, and Red Hat OpenShift use container technology to allow apps to span your private cloud and an array of public clouds. And Microsoft Azure Stack* extends its cloud right onto your premises using the same cloud application platform used in its public offering.
What is clear is that the rate of innovation in the cloud is quickening, and it continues to evolve and transform. It’s becoming a platform for analytics, artificial intelligence, high-performance computing, and network transformation.
The majority of today’s modern cloud data centers—public and private—are powered by Intel® Xeon® processors, and our latest offering, the Intel Xeon Scalable Processor family, builds on that legacy and delivers new features for the cloud—helping you deliver secure and agile digital services. It enables a common processor architecture spanning low-power, 8-core systems for edge computing up to an 8-socket system comprising 224 cores and supporting 12 TB of memory—all optimized for compute, storage, or networking workloads. It’s one of the ways we’re working to help you on your journey to the transformed computing environment that will enable you to bring your own innovations to the market faster and cheaper.
By William (Bill) Giard, CTO, IT Transformation, Data Center Group, Intel