Move Faster, Decrease Risk and Lower Operating Expense with Integrated IT Systems
Integrated IT systems, also known as converged infrastructure (CI), is often perceived differently by many people. However, most would agree that CI is fundamentally the collocation of compute, storage and network resources, often into one or a small handful of cabinets, managed as a single IT asset. Whether pre-engineered and defined on paper by a reference design, or actually prefabricated by one of several turnkey vendors, these converged stacks are typically characterized by a high degree of virtualization and automated software management. Virtualization enables pooling of compute, input/output and storage resources. Pre-engineered and tested logic gives rise to automation schemes that make the rapid configuration and allocation of these resources possible for a wide variety of workloads. This automated flexibility lends itself well to self-service and “pay as you go” models typical of private cloud infrastructures. Cisco’s highly successful Unified Computing System (UCS) is a leading example of this integrated architecture.
What really catches my attention as a “power, cooling, and rack guy” is the pre-engineered and often prefabricated nature of these systems. This is what makes the concept of converged infrastructure so interesting. Perhaps the most impactful benefit of CI is the ability to provision, test and deploy IT faster, as well as a major reduction of time and manpower needed to manage and maintain the infrastructure once operational. These valuable benefits are a direct result of the fact that all components and sub-systems are designed, tested and supported together as a system by the vendors or system integrators. However, these benefits are more than just vendor claims; they are increasingly proven and being documented by third parties and real customers. Case studies show in dramatic fashion how and why CI is pushing the industry further down the path of standardization, modularity and prefabricated construction.
I have been struck by the clear parallels of CI to the physical infrastructure of power, cooling and rack systems. Schneider Electric and other vendors have been offering pre-engineered, modular and prefabricated solutions for years. The benefits are virtually the same as those for CI: faster deployment, easier configuration and maintenance, increased flexibility and improved efficiency. Just as with traditional IT gear, an integrated system needs adequate power, cooling, physical security, connectivity resources and a place to rack them. However, because CI solutions are, to a large degree, pre-engineered and documented, specific, pre-configured solutions are faster and easier to get “off-the-shelf.” This is where physical infrastructure vendors can really shine. CI is fundamentally about speed, efficiency and agility – keeping IT aligned to the business. It’s up to vendors and integrators to help make that happen by providing physical infrastructure solutions that are easy to select, order, deploy and maintain.
One way vendors are doing this is by offering pre-configured solutions that combine racks, rack PDU strips, battery backup power, environmental monitoring, physical security, air containment, DCIM software and other infrastructure. These solutions can be user (or partner) configured easily through online design tools that match specific converged stack requirements. In some cases, pre-configured solutions exist “off-the-shelf” for well-defined stacks, such as VCE’s vBlock systems. These offers are pre-engineered to work and fit the specific stack, meeting the required power ratings, dimensions, number and type of connectors and so on. These “off-the-shelf” solutions make specifying and procuring power, cooling and infrastructure management software much simpler and faster.
Reference designs are another method vendors are using to simplify the data center. Data center reference designs are documented and validated conceptual plans detailing how physical infrastructure systems are configured, connected and laid out for a given set of performance requirements. Their standardized performance attributes, equipment lists, floor layouts, and “one line” diagrams make it easy for project teams to rapidly compare and contrast design alternatives. Validated designs allow architects and design firms to agree on key project parameters (such as IT capacity, density, redundancy, PUE and budget) earlier during the programming phase of a project. Early planning is critical for ensuring the rest of the project phases go well.
Implementing prefabricated data center power and cooling modules is another strategy that simplifies and accelerates data center projects. Similar to converged IT infrastructure solutions, prefabricated power and cooling modules are pre-manufactured and integrate otherwise complex and disparate sub-systems, thereby increasing predictability and reducing much of the custom on-site engineering and integration work that is common with traditional stick-builds and projects.
Finally, lifecycle services such as facility management, energy assessments, product training, maintenance, vendor management, and installation and commissioning can fill knowledge and resource gaps that would otherwise slow growth or reduce responsiveness to changing business requirements.
The success of integrated IT systems is accelerating the rate of adoption of pre-engineered, standardized, and (often) prefabricated systems in the data center industry. A real improvement in the responsiveness and agility of IT, as well as a general reduction in overall total cost of ownership (TCO), is driving this trend. Pre-configured solutions, prefabricated modules, reference designs and lifecycle services all strive to achieve the same goals of converged infrastructure solutions—increasing agility, speed, simplicity, and resiliency at a lower operating expense.