Enzo Greco, chief strategy officer at Nlyte Software, outlines some simple steps to ensure the efficiency of your critical infrastructure.
Efficiency is a critical metric for any organisation. For data centres, it is often defining, as efficiency is central to their operating model. We are able to measure inputs, processes and outputs with incredible ease, with sensors, machine learning, and dashboards making the manual labour and guesswork of a few short years ago barely a bad memory.
Now workflow, analytics, a building-to-rack approach, remote management, and holistic capacity management all align and feed into a corporate awareness that allows facilities, infrastructure, and business to operate as a truly seamless and sleek organism. But this only happens when the business is in sync with itself, its processes and people pulling together and working with consistent, end-to-end information with clear goals.
Here are five steps that any organisation can begin putting into practice today to become infrastructure efficient and ready to be as dynamic as the business and market dictates.
1. Acquire an end-to-end workflow and understand all points on the journey
Traditionally, Data Centre Infrastructure Management (DCIM) and Building Management Systems (BMS) have been silos with distinct workflow systems. But there are significant benefits to these systems being integrated. Integration promotes efficiency and visibility. For example, when the two systems are aware of each other it enables key functions such as:
- Automating failure recovery, especially in remote facilities where we don’t have the luxury of “walking the halls”
- Providing accurate and automated provisioning for everything from the critical infrastructure (power, cooling, space) to the virtualised applications themselves
- Providing holistic alarm and alert management – making it intelligent and end-to-end: Wherever an event arises, it is important to correlate all the different alarms and alerts to provide context and identify the root cause as quickly and accurately as possible
- Managing service requests holistically. For example, what workloads are at risk when a chiller undergoes maintenance, either planned or unplanned? This is a straightforward yet common use case that exemplifies the requirements and benefits of end-to-end visibility and workflow.
2. Gain richer analytics to understand more of your universe
As any data centre manager and head of computing infrastructure knows, the scale and complexity of a modern data centre necessitates rich, multivariate analytics from a variety of sources: Virtual, edge, colocation, and including both BMS and DCIM.
All this information (and there is a tremendous amount) should be incorporated in order to manage processes, discern patterns, analyse possible outcomes, increase efficiency and head off problems before they happen.
There simply are too many inputs with too many possible ideal outcomes to manage data centres traditionally; analytics are ideally and uniquely suited to addressing this challenge and opportunity.
3. Take a holistic ‘building-to-rack’ approach
With common visualisation and consistent terminology of everything from buildings to racks to workloads, and everything in between, capacity management and accurate reporting are simplified, and systems are easier to use for all concerned.
The value to business are many: Increased availability, greater optimisation and efficiency, and the ability to make better management decisions. All-in-all, taking the zoomed-out to zoomed-in approach enables the business to see all things in the right place and to the right plan when it comes to measuring and operating the facility best.
4. Enjoy remote management
The outsized growth of edge computing (growing far quicker than any other infrastructure segment) is driving many key requirements, none more so than remote visibility and management. Smaller, dispersed facilities are often unmanned – so a manager does not have the luxury of being on-premise to solve for problems then and there.
Automation is vital for smooth-running facilities under a wide variety of expected and unexpected scenarios, with BMS and DCIM integration being required. For example, holistic alarm and alert management: If a manager isn’t there, they’ve got to be able to discover what’s going on at the remote facility in near real-time to avoid disaster and speed recovery. Further, edge computing is driving analytics as a way of “self-learning” the numerous issues that may arise.
5. Create holistic capacity management
You can’t do a good job with capacity planning if you’re just looking at the facility and not the applications, users, the demand, and so forth. Dynamic environments demand visibility into what it costs to run a workload on-premise, in the cloud, and in a colocation facility, so decisions can be made for the best optimisation of resources.
Complete information across both BMS and DCIM is required for determining the best decisions for commissioning and decommissioning, scheduled maintenance, inventory control, and power and thermal changes.
An underlying requirement for the above is for the organisation to have an understanding of their assets, both physical and virtual: What they are, where they are, their current state and the context around them.
Consider the simple example of servers: How many are there? Where are they? How much free capacity do they have? Do they have the right versions of software to run that shiny new ERP system? Are they “secure”, in the many ways we can define that?
Asset management becomes ever more important with the growth of edge computing, as this is a critical, growing, yet often “hidden” part of your computing fabric. The first step in any journey should be to discover and categorise the numerous assets every organisation has.
A carefully considered infrastructure efficiency plan will increasingly have automation and intelligence built in so that critical infrastructure never falls into a critically-endangered state that risks imperilling business systems or customer service delivery.
Far from being a concern to only the key critical infrastructure or data centre team, an understanding of how to drive efficiencies should in fact now be on the agenda for the senior leadership team given the criticality of technology infrastructure to every facet of business delivery in the information age.
It all starts with simple steps. Even if a technology partner is required to implement the right process, it’s now within the capabilities of non-specialists to manage infrastructure assets and maintain optimal efficiency given the sophisticated tools available to large enterprises and data centre service consumers alike.