Data centre managers are constantly seeking more productive ways of managing their operations and with automation tools like Artificial Intelligence coming into play, they are able to do just that. Here, Herman Chan, President, Sunbird Software, talks to us about some of the key benefits of leveraging automation to get ahead in the data centre space while driving innovation in their operations.
Why are data centre sites becoming increasingly complex and difficult to manage?
The demand for data centre services has never been higher. Today’s data centre managers face skyrocketing data storage needs, increasing rack densities, more connected devices and massive energy consumption. The demand for data centre services is driving more technology deployment and faster upgrades of outdated power-hungry systems and components, in turn driving more change within the data centre.
To add to this complexity, modern data centre environments are distributed across many remote sites such as colocation and Edge deployments.
Data centre managers struggle to remotely track all their assets, plan and manage their infrastructure capacity, perform changes at speed, increase energy efficiency and safeguard their assets and data from physical and cybersecurity threats, all while minimising risk and maximising uptime.
To make matters more difficult, many data centre managers are forced to work with a myriad of disparate tools that lack integration. They have siloed databases with no single source of truth and must sacrifice productivity and data accuracy by manually entering data into multiple systems.
Why — and how — should data centre managers follow best practices to maintain uptime, increase efficiencies and improve productivity?
Maintaining uptime, increasing efficiency and improving productivity are arguably the three most important objectives for any data centre manager.
Downtime has been estimated to cost US$5,600 per minute on average and that figure can be much higher for larger organisations. Downtime leads to lost sales, damaged brand reputations and SLA payouts.
Driving efficiency is more important than ever as most data centre managers now have corporate sustainability initiatives they must comply with. Increased efficiency also leads to reducing operating costs and extracting more value out of existing capacity.
By improving productivity, data centre managers can devote more resources to strategic projects rather than manual, time-consuming tasks. This can result in getting services online faster and responding to outages quicker.
Over the years, there has been plenty of new insights and process innovation that are driving data centre management best practices. While there are many, the ultimate data centre management best practice is ‘integration with automation’ or stitching together the multitude of data centre operations tools and having a centralised view of all resources and capacities from integration of CMDBs, DCIM, BMS, ticketing and other tools. This real-time sharing of salient data makes data centre asset, capacity and change management easier and faster than ever before.
What trends have you seen regarding data centre managers operating remotely and what challenges does this pose?
Remote data centre management is the new normal. Whether it’s due to COVID-19 and work from home, security requirements to minimise onsite staff, or the distributed nature of modern data centre environments, most data centre professionals are no longer working onsite.
When working remotely without the right tools, data centre professionals struggle to achieve their core objectives of maintaining uptime, increasing the efficiency of capacity utilisation and improving productivity.
Maintaining uptime remotely is a challenge when you don’t have information on the real-time power and environmental conditions within the data centre and you can’t go onsite to take manual readings. A circuit breaker might trip or hot spots might damage equipment before you even know an issue is developing. When there is a serious problem, it can take more time to troubleshoot and resolve it.
Capacity planning is difficult because there is no visibility into how much space, power, cooling and port capacity is available. Remote data centre managers don’t know where the best place to deploy new equipment is and manual spreadsheets and diagrams are error-prone and difficult to manage.
Finally, improving the productivity of people is arduous because remote teams lack tools that drive collaboration and data sharing. This leads to data siloes, multiple manual data entries and inaccurate data that can cause larger problems like downtime and stranded capacity.
What role does automation play in addressing the above challenges?
Automation can help alleviate many of the pain points of modern data centre management.
Maintaining uptime is easier with the aid of automatic alerts of potential power and environmental issues. Data centre managers can set warning and critical thresholds on the live measured readings from power meters and environmental sensors. Then, upon a threshold violation, you will be the first to know of conditions that may cause downtime so you can proactively remediate the issue before there is a serious problem.
By the same token, you can increase energy efficiency by being automatically notified when temperatures reach your lower thresholds, signifying that you are overcooling and wasting energy. Then, you can raise your temperature set points to save energy and money.
Finally, automation can dramatically improve productivity by saving time and improving data accuracy. Integrating data centre management tools reduces manual data entry and breaks down information silos, giving data centre teams more time to work on strategic projects, get services online faster and collaborate around shared data to drive better decisions.
In what ways can data centre managers introduce automation to their management practice?
Our customers are the most modern data centre managers in the world and we’ve been fortunate to witness first-hand how they are introducing automation into their management practice.
At a recent customer event, a customer presenter early on in their automation journey said, ‘What we’re trying to do as an organisation is to get out of the data world and into the information world’.
The way they were achieving that was to make DCIM software ‘the centre of the universe for the data centre’.
That is the story we hear from customers who are successfully implementing automation.
With a second-generation DCIM tool that has fully documented and open APIs, data centre managers can easily integrate with their CMDBs, ticketing systems, BMS and many other data centre management tools. Then, they can automate virtually anything they desire such as provisioning and orchestration, virtual machine management, parts management and decommissioning. They can also enjoy all the built-in automation capabilities within their DCIM tool such as thresholds and alerts and scheduled email of charts and reports.
How can automation assist with device power budgeting?
First, consider the traditional approach to device power budgeting and rack power capacity planning. Data centre professionals often derate the server nameplate value to 60% or 70% to allow for a safety margin. This process is manual, estimated and largely inaccurate. The result is excessive stranded capacity.
This process can be entirely automated with our patented Auto Power Budget feature. Auto Power Budget automatically calculates an accurate power budget number for each make and model instance of a device based upon the actual measured load of that device in your environment running your applications. Power budget numbers are automatically updated every week to ensure a high degree of accuracy.
With this feature, Comcast was able to get 40% more utilisation out of their rack power resources and eBay saved US$120,000 in a single project by reducing the number of cabinets needed to deploy new services.
What trends do you expect to see looking ahead when it comes to automation in data centres and how can organisations prepare in order to stay ahead?
I expect to see more adoption of data centre management tools that are open and ready for integration with out-of-the-box connectors, along with improving process maturity to enable forward-thinking data centre managers to drive automation in their operations.
By integrating systems and leveraging automation capabilities in their tools, our customers are saving time and dramatically simplifying data centre management. The best practices they are pioneering are getting noticed by other data centre professionals.
To prepare, organisations should begin researching how these leaders are leveraging automation to enable a single source of truth, eliminate human error, improve collaboration, increase the utilisation of existing resources, maintain uptime, improve workflow and productivity, and more.
To stay ahead, organisations need to understand which of their unique pain points can be resolved by automation and deploy modern data centre management software that will help them get there.Click below to share this article