Kevin O'Gilvie, Director, IT Infrastructure Controls, Columbia University in the City of New YorkIn today's technology environment, whether you are working for a mid-sized or large organization, managing various components of a data center can be a daunting and expensive task. There is a strong demand for simplifying data center management in order to increase productivity, streamline processes and reduce costs.
Traditionally, IT organizations manage their legacy environments as multiple separate silos. Servers, storage, and network resources are managed on separate devices. This approach slows down service provisioning and makes it difficult to adapt quickly to changes in service workloads. In order to facilitate tasks such as upgrades, planned migrations, and maintenance, a combination of resources from various systems and infrastructure groups is usually required. Additionally, the siloed approach leads to the underutilization of available capacity, because resources cannot readily be applied to any given workload.
"IT organizations need to be able to provide efficient and agile data center solutions in a responsive manner that is cost effective and resource efficient"
The idea of throwing money at technology without having a business use case is no longer sustainable. Various technology trends come and go. One thing that has always remained the same is that businesses need to cut costs and do more with less. In order to keep up with the changing business demands, some companies are abandoning the traditional IT model to go with public cloud offerings. Although it is a viable solution that can offer reduced costs and accelerated deployment, public cloud faces compliance and IT governance issues that limit which IT services can be migrated to it. IT organizations need to be able to provide efficient and agile data center solutions in a responsive manner that is cost effective and resource efficient.
Converged infrastructure (CI) is an approach to data center management that aims to minimize compatibility issues between storage systems, servers, and network devices. In addition, CI also seeks to reduce costs that are associated with floor space, cooling, cabling, and power. CI combines storage and computation into a single solution that is dynamic and powerful. CI also consolidates the components of the infrastructure, simplifying management and change. It can be beneficial to organizations looking to reduce costs of deployment, simplify management and reduce maintenance and support costs.
There are many CI solutions in the market. Selecting the right CI product for your environment requires proper research, as well as performance and capacity planning. The most important factor in your decision should always be workload. The workload is the amount of processing that the computer system must complete. It is critical that you work with application and systems owners to determine performance characteristics of your service workloads, as well as the capacity requirements before making the final decision. Every application or user session that interacts with the application can impact workload or storage. Factors such as latency, storage replication, virtual machines and input/output operations per second (IOPS) are all key to determining whether a given workload is ideal for the chosen solution.
CI by design allows focus to be put on the workload, not the infrastructure components. Components in a CI platform are fully integrated and need to be configured to match your workload and storage demands. This will allow focus to be directed to virtual servers, operating systems, applications and security without worrying about the physical infrastructure.
CI does have its limitations and is not a magic pill for every environment. CI provides limited flexibility as it is a fixed solution. If you are looking for more elasticity, hyper-converged infrastructure (HCI) would be a better solution for your environment.
HCI provides many of the same benefits of CI and at the same time addresses some of the limitations. While CI combines existing storage, compute and network into a validated configuration, HCI merges the storage and computation into one server. Hyper-converged infrastructure utilizes multiple servers that are clustered to provide a platform for the workload and storage. Whereas with CI scaling would mean purchasing a new CI product, hyper-converged infrastructure allows scaling by adding new servers to the cluster, which is a much smaller purchase than a new CI unit. The cluster can easily scale out as more servers are added, and any number of servers can be added to accommodate the changing workload. The difference between CI and hyper-convergence is that hyper-convergence utilizes emerging hardware and software platforms, utilized in a new operating model. Hyper-converged infrastructure utilizes existing and proven hardware in a new way.
Convergence, whether CI or HCI, enables the pooled management of distributed resources to be automated, which eliminates the cost, delays and potential errors associated with current manual processes. This automation allows fast and easy provisioning of new applications or services. It also allows the IT environment to address workload bottlenecks automatically — whether that bottleneck is in CPU cycles, server memory, storage I/O or network throughput. Just as important, automation allows resources to be released when they are no longer needed. Through this combination of pooled resources and better automation, convergence empowers IT to define policies that govern the relationships between infrastructure resources and business workloads. These policies can be driven by business priorities, service level objectives, compliance requirements and other relevant parameters. So, in addition to streamlining operations, IT can leverage convergence to bring coherent governance to its day-to-day operations.
