Digital Next Advisors and HPE Pointnext Services Advisory and Professional Services experts map technology and offer best-practice advice to help you: deliver on digital ambition, create new value streams, speed up innovation, improve customer experience, and more.
BrandPosts are written and edited by members of our sponsor community. BrandPosts create an opportunity for an individual sponsor to provide insight and commentary from their point-of-view directly to our audience. The editorial team does not participate in the writing or editing of BrandPosts.
By Munther Salim
A data center, wherever it is, owned by you or a cloud provider, is the foundation upon which all future digital transformation is built. In choosing your IT supply path, most start with a clear understanding of current and future needs, working with trusted advisers to arrive at the best solution. That’s the right way to approach any kind of transformation. Businesses today can’t afford to be tied to legacy infrastructure. That’s what has driven cloud migration. But performance efficiency begins in the facility itself.
Mostly, “software-defined data center” (aka SDDC) is interpreted as “IT.” Data center facilities engineers are used to this, but would insist that software-defined begins in the brick and mortar. Once that is understood it means the CIO will eventually succeed in providing all functionality to support the company’s IT transformation strategy and roadmap.
Data Center Management Services (DCMS), as applied to the building, is a straightforward implementation and integration of software suites, processes, procedures, and organization that will provide the foundations of software-defined layers.
Having a software-defined data center enables you to measure and manage data about all the active components in a data center, end-to-end, including IT, network, applications, and facilities. The real mistake is to consider either IT infrastructure or facilities infrastructure as separate elements. Building the bridge between IT and facilities is the answer, and it involves shortening the height of several management layers.
A proper DCMS approach is an SDDC foundation. It comprises four layers, each one addressed through tools, processes, procedures, and organization components. (figure 1)
Layer 1 is almost always available: Supervision and data collection
Even if not tuned for DCMS, this layer is about data center supervision and monitoring tools as Building Management System for facilities, or IT supervision suites. At this same level is dedicated software that bring intelligence at the distribution layers as intelligent rack PDU systems, or intelligence network cabling systems. This layer also includes the supervision of middleware made up of virtualization, containerization, network management, and database management.
Layer 2: Operations management
Successful enterprise operations have efficiently run data centers that deliver the right service level and capacity at the right time, and at a reasonable cost. Without strong knowledge, processes, and operational know-how from trained staff, and with constantly changing processing environments, the large-scale capital that enterprises invest in a data center likely won’t yield a satisfying operational performance.
HPE is unique as a technology vendor in that its solutions incorporate both compute and facilities technologies and practices. It’s positioned to help organizations develop and maintain well-run data center operations to meet service-level objectives. It offers governance rules, operational intelligence and procedures, runbook-based documentation, staff selection guide and training, and performance measurement indicators for new and existing data centers. HPE adopts a systematic and multistep approach to assess and evaluate the existing data center operation status, ultimately documenting operations into the “DC runbook.”
Layer 3: DCIM as a key element
The third and key layer is the Data Center Infrastructure Management (DCIM) layer. DCIM systems bridge the gap between IT and facility management by creating an IT and facility asset database which is directly linked to space, power, and cooling on the facility side and compute, storage, and networking on the IT side. This allows the system to optimize provisioned infrastructure, accurately and quickly deploy new systems with related power and network connectivity, and perform accurate capacity planning.
The new paradigm involves managing the data center not only for reliability but also for efficiency. It is challenging to manage what is not measured. Reliability and efficiency of the data center cannot be sustained without an IT-facility integration tool such as DCIM.
DCIM enables energy efficiency because it is a holistic solution that tracks and manages both facility- and IT-based assets. It also automates the planning for new assets and removing old assets. Without it, data center managers have to manually query available resources such as rack space, power capacity, network ports, and patch panel ports. This process is labor intensive with real costs. Running automatic modeling analyzes various scenarios to configure the data center and provide real-time outcomes for making intelligent decisions around optimizing the data center.
To enable energy efficiency, static data must be collected, as well as continuous and granular dynamic data from the various data center assets. The DCIM dashboard provides energy benchmarking between multiple data centers. This allows operators to compare energy metrics and understand both challenges and best practices at their various sites. It also allow operators to strategically plan capacity upgrades and IT changes with overall energy metrics in mind, including decisions for colocation and cloud deployment.
Layer 4: Data center efficiency and service portal
This fourth layer will complete the picture for a SDDC-ready system: Global data center efficiency is measured, managed, and reported on top of the first three layers of this model.
This comes with inventory and asset management, configuration management, capacity management, integration management, intelligent dashboards and reports, all of which together provide the right information for the right decision to the right stakeholder, as shown in figure 2.
Having a Service Portal that will publish data and gather user requests without an integrated DCIM system leads to the miss-use or “no-use” of the capability of such a system.
Integrating Data Center Infrastructure Management into software-defined data center policies has only been mainstream for about 10 years. Previously, IT management and facilities management would often butt heads, mainly because of disagreement on domain and perceived speed of response to functional needs. Integrating DCIM has helped facilities and IT teams to work together against a common dataset so that each can be better informed.
By plugging DCIM tools into building information systems, data center managers can understand the constraints that are outside of the data center itself. Likewise, the facilities team gains greater insights into what will be required from the facility when it comes to power and cooling and can start to plan accordingly.
Most organizations are still allocating the bulk of the IT budget toward “keeping the lights on.” It’s tough to transform your business when you are battling to handle and keep legacy IT architectures, and even tougher if you do not modernize the data center itself.
Dr. Munther Salim is a Distinguished Technologist, HPE Pointnext Services consulting. His focus includes data center green design, energy efficiency and sustainability, and data center facility operations consulting. He is an active participant and author in many industry organizations such as the Green Grid, 7×24 exchange, and ASHRAE. He has also advised a number of governments and governmental advisory bodies as they determine their data center energy efficiency strategies in the context of global targets.