- Senior Director of Facilities »
- Director of Facilities and Fleet Management »
- DIRECTOR OF COLLEGE FACILITIES »
- ELECTRICIAN »
- Construction engineer, U.S. Dept. of State »
7 Steps to Data Center Modernization
September 22, 2017 - Data Centers
By Mark Gaydos
“Life is never stagnation. It is constant movement, unrhythmic movement … constant change. Things live by moving and gain strength as they go.” — Bruce Lee
Change is difficult. In an environment as complex and constantly in flux as a data center in an institutional and commercial facility, the saying “Change is difficult” is especially true. In addition, maintenance and engineering managers also tend to adopt another mantra “If it isn’t broken, don’t fix it.” But this mind frame outlives its usefulness quickly.
Modernization in a data center is absolutely vital. Data centers must stay current so as not to run the risk of falling behind technologically and ceasing to remain competitive. Without updates to security, a facility is vulnerable to new, more complex cyberattacks. Based on most recent cyberattacks, there is no doubt that hackers who strive to penetrate the data center infrastructure are the perpetrators and creators of extremely sophisticated malware.
In addition to the security issues, changing demands on the data center are forcing managers to do more with less. This means they must squeeze every possible ounce of capacity out of the infrastructure while delivering applications and information technology (IT) at an accelerated pace. At the same time, the goal is five 9s of reliability, all while not incurring any additional risks.
There are ongoing inhibitors to data center evolution. Almost 80 percent of chief information officers (CIOs) are hindered by legacy infrastructure, according to a recent survey.
Yesterday’s infrastructure was not built to support today’s instantaneous consumer demands. New challenges like the exponential proliferation of data, Big Data, Internet of Things (IoT), a mobile workforce, rising energy costs, and the balancing act of selecting a hybrid — on premises, off premises, cloud, edge, etc. — data center strategy simply cannot be met with aging infrastructure. With modernization comes the agility that enables innovation so the facility can respond more quickly to demands. Staying current on new technology innovations helps boost company revenue, cut costs, and even rightsize infrastructure for efficiency and reliability.
Data center modernization makes expectations easier to achieve because managers can take advantage of new technology to fulfill them. In fact, data center automation puts the entire organization on the same page and reduces siloing, errors and extraneous manual labor while delivering a high return on investment (ROI) and cost savings. Newer equipment is easier and less expensive to operate, uses less energy, and is less likely to break down and cause an outage. Up-to-date IT creates greater opportunities to save money, reduce risk and provides customers a better, more individualized, and personalized experience.
If organizations balk at the cost of modernizing, they also should think about the much higher cost equated with outdated, legacy IT equipment.
To effectively modernize a data center, it takes some upfront work, time and expense, but the long-term payoff is well worth the effort. Implementing these seven steps to modernize and update infrastructure is well worth considering for any data center to keep customer data safe and secure in a resilient environment.
Programmatic approach to tech refresh. The ability to be logical and vigilant about tech refresh — retiring and replacing legacy equipment — depends on keeping track of every asset in the data center through its entire lifespan. This is sometimes referred to as dock-to-decom tracking an asset from the time it is received on the loading dock — or even before, when it is requisitioned or purchased — to the day it is decommissioned. This way, management knows the exact age of each old piece of equipment, where it is in its warranty cycle, and when it is at the end of its useful life.
Being systematic about when a piece of equipment becomes a good candidate for retirement is beneficial because older equipment generally costs more to maintain and operate. In addition, older devices are less productive, consume more power and are prone to failure. There comes a point in an asset’s lifecycle when it becomes more expensive to maintain than to replace.
Data center infrastructure management (DCIM) software helps data center personnel keep track of IT assets by knowing components that are coming up for retirement and the level of energy a particular piece of equipment is using, then sending an alarm and alert warnings before a malfunction.
Workflow process management. DCIM workflow tracks everything done to every asset – the work performed, the person who did it, when they did it, etc. By having one central repository for this information, resources are scheduled more easily, work orders are generated, and automation ensures that there are no knowledge gaps in the team. Workflow increases productivity by helping teams work more efficiently by adding process consistency and accountability.
Complete power chain hardening with power failure simulation. The U.S. economy loses up to $550 billion annually due to power outages and other disturbances, according to Hexa Research. Ensure your data center’s power source is hardened and disaster-tested.
This process allows a manager to understand the complete stack, from the data center floor up through the application, as well as map out what power systems support which assets and apps so the team is aware of what might be at risk during a power outage.
Additionally, it allows managers to see who has access to the power systems, as well as when and where the passwords last changed.
DCIM software should have the ability to perform virtual power failure simulation so the data center team can pinpoint what would happen to the critical infrastructure if a particular system or piece of equipment failed. It also should provide information on when the last actual, controlled failure test was performed and whether power failure recovery is part of the business continuity.
Tight integration with ITSM process management. Connecting the DCIM solution to information technology service management (ITSM) solutions is a vital step in a data center modernization plan. ITSM systems, such as BMC, HPE and ServiceNow, should be integrated with the data center systems and share information back and forth. Changes originating in the ITSM change management system must be passed to the data center for execution. This provides an important link between facilities and IT personnel. Prebuilt connectors between the ITSM and DCIM systems simplify integration greatly.
Hybrid strategy that incorporates on-premise, colocation and public cloud. The future of data center strategy is hybrid. The strength of this approach lies in flexibility and adaptability. With a hybrid strategy, applications and workloads can run where they make the most sense.
Hybrid provides a choice when it comes to the most ideal computing environment – using a mix of on-premises, private cloud as well as a third-party, public cloud provider services with interoperability between the two platforms. A hybrid cloud also allows a range of flexibility for businesses as computing and budgetary constraints change, allowing workloads to fluctuate between private and public clouds. This also allows businesses the option to deploy far greater data when needed.
Workload segmentation. Flexibility is the ultimate goal here. Selecting the best place to run the workloads is really dependent on an organization’s priorities for those workloads. An organization needs to address these issues:
• Is the concern about cost-effectiveness?
• Is performance important to this particular workload?
• Does this data belong in an “edge” facility to keep the computing and analysis closer to the end-user to cut down on latency and deliver a better customer experience?
• Is the workload something the management would rather keep close and in-house, for security, compliance or other reasons?
Hybrid strategy enables the flexibility to move workloads around to achieve goals.
Physical integration with virtualization/container layer to reduce workload risk. Virtualization definitely saves on resources and is a wiser use of capacity. Virtualization allows an organization to create a virtual version of computer network resources, hardware platforms, storage devices, etc. With server virtualization, software actually mimics hardware – like a CPU’s memory or network traffic or others. Obviously, software performance levels in a virtualized system do not match the hardware levels, but they allow for greater flexibility and control since users do not need to use all the hardware capacity.
But with virtualization, often the user does not even know the location of the physical layer and is not privy to the physical address. With DCIM’s integration into the virtualization layer, the user has a holistic view of what is running on each of the virtual servers, and which physical hardware the workloads and applications are running on. This visibility enables management to see where the most important workloads are running so they can protect those specific physical servers.
Modernizing the data center is not a matter of if. It’s a matter of when. The data center’s facility infrastructure must be kept up to date, or it will lag behind. A modernized data center environment will be more secure, affordable and efficient not to mention more adaptable to new data challenges and business opportunities. It will deliver the customer experience that both internal and external customers demand.
A DCIM solution remains at the core of data center modernization efforts. With DCIM, facility managers can easily keep track of assets throughout their useful lifecycle, safeguard and track power for efficiency improvements, deliver workflow so all internal teams can work together more effectively, help distribute workload most efficiently and integrate with other systems, even with the virtual layer.
Without a DCIM deployment, a data center cannot be considered a truly modern facility capable of handling the business needs of the ever-changing technology ecosystem.
Mark Gaydos is chief marketing officer with Nlyte Software, the leading data center infrastructure management (DCIM) solution provider for automating the management of infrastructure in data center or colocation facilities.