Data Center of the Future
By Ari Bose, CIO, Brocade
Today, most enterprises have heterogeneous IT footprints with both on and off-premise assets, which can be difficult to manage since there is no “single pane of glass” that provides visibility into the performance of all assets. Today, there are multiple alternatives to host applications on and off-premise, but the right approach isn’t evident. While the trend is to host applications off-premise,78 percent of IT leaders believe that some aspects of their application resources will remain on-premise for the foreseeable future. At the same time, IT leaders want to transform the data center to enable more nimble operations and increased innovation. A hybrid on and off-premise model with a cloud management platform which enables orchestration and management of services would enable organizations to become more agile and responsive today. What are the key next steps to achieve this model?
Step 1: Define which portion of the data center remains on-premise and which portion is housed in the cloud. Below are three ways to categorize business applications to determine the optimal data center for the business:
1. On-Premise: Major regulatory and compliance applications (e.g., to enable HIPPA and SOX compliance) with a clear focus on high availability and operational efficiency should continue to run on-premise with highly virtualized clustered instances, which can be upgraded in-line and have built in disaster recovery capabilities. Ethernet and IP Fabric for on-premise data centers provide the capabilities to deliver a zero down time infrastructure while enabling in-line upgrades.
IT leaders want to transform the data center to enable more nimble operations and increased innovation
While cloud-based solutions exist, if privacy, security, and regulatory risks are salient concerns, enterprises should house these applications on-premise in the near term.
2. Infrastructure as a Service (IaaS): Innovation and analytics applications with high computing needs, storage elasticity, and burst capacity should be run in a private cloud, as they require environments, which support Big Data to enable robust analytics. Examples of these are machine learning, factory data analytics and applications that support and manage IoT.
3. Platform as a Service (PaaS): SaaS applications with proprietary architectures (e.g., cloud-based ERP and CMR systems) must be housed in a public cloud. Other applications which enable faster time to market or help provide a competitive advantage can be run either on-premises or hosted in a private or public cloud. The choice of environment should depend on latency requirements, unique processes, service-level requirements, security requirements and concerns, and desired scalability.
Step 2: Assess what type of “single pane of glass” is required. Today, most organizations have multiple dashboards to manage their heterogeneous assets. IT teams are forced to “stitch things together” to provide an integrated view of on and off-premise asset performance. However, as IT organizations become more business-driven, the dashboard they leverage needs to report on the availability of services, not assets. A single cloud-based dashboard would enable enterprises to monitor and manage services provided by all applications. The onus is on IT leadership to define which services require reporting, and on our industry to help develop this “single pane of glass.”
While defining and adopting the “right” technologies is a step in the right direction, it isn’t enough. An enterprise’s data center strategy must consider the skills and capabilities required to maintain and evolve the data center. The shift in mindset from boxes to services demands IT resources who understand they develop and deploy business services, not IT assets. These individuals must be able to understand the “big picture” and address abstract problems in order to deliver business value and facilitate the implementation of the organization’s strategy. IT organizations, which only focus on technical competence without linking this competence to the business will struggle to truly transform. As we all know too well, change isn’t easy; IT organizations should assess which skills and capabilities are required, identify gaps, and begin addressing those gaps today.
Curating the right distribution of on and off-premise services that matches business demands will allow IT to become a true enabler and strategic partner to the business. Integrating a hybrid model to transform decades-old IT footprints and developing the required skillsets today will provide the foundation an enterprise of the future needs to respond to market needs and increase innovation.