Running a data center is a complex undertaking. In addition to maintaining the strict physical security measures and logical security protocols needed to secure customer data, facility personnel face an ongoing challenge of optimizing IT infrastructure to improve power efficiencies and maximize cooling capacity. Quality data center operations are a key differentiator for colocation customers and managed service providers (MSPs) looking for the best possible partner to house and manage their IT infrastructure solutions. This makes it critically important for a facility to implement a variety of best practices to improve its data center operations.
Storage technology has come so far and so quickly that many people don’t even stop to think about storage capacity anymore. With the price of solid-state drives (SSDs) continuing to fall and new forms of magnetic disc storage becoming more readily available, companies have more than enough space to store the massive amounts of data they’re collecting from their networks and internet of things (IoT) devices. This unstructured data serves as the fuel for powerful analytics programs that churn through information looking for trends and actionable insights that can inform decision making.
But what about all the data that doesn’t get used? According to a 2016 study, redundant, obsolete, and trivial (ROT) data constitutes about 33 percent of all data stored and processed by organizations. Left unattended, that data could cost upwards of $3 trillion globally to manage by 2020. By implementing good data hygiene, data centers can help their customers better manage their ROT data and free up valuable computing and storage resources for more critical needs. Reducing the burden on IT infrastructure can have beneficial impacts on data center standards because it means computing equipment will have lower power and cooling demands. It also allows the facility to allocate its storage and processing resources more effectively.
Few things are more critical to data center operations best practices than an effective data center infrastructure management (DCIM) platform. Managing a data center without DCIM software is like trying to sail a boat in complete darkness. It’s nearly impossible to know what’s happening in the moment and even minor problems can be extremely disruptive because they take the facility by surprise.
Implementing DCIM tools provides complete visibility into the facility’s IT infrastructure, allowing data center personnel to monitor power usage, cooling needs, and traffic demands in real time. They can also analyze historical trends to optimize deployments for better performance. With a DCIM platform in place, IT support tickets can be resolved quickly and customers can communicate their deployment needs without having to go through a complicated request process.
Deployments matter, especially when it comes to issues of power distribution and rack density. Inefficient deployments can lead to problems like wasted energy going to underutilized servers or too much heat being generated for the cooling infrastructure to manage. The layout of the data floor can be subject to quite a bit of change, especially in a colocation facility where new servers are being deployed on a regular basis. Data centers need to be aware of how every piece of equipment on the data floor interacts with the others in order to optimize the environment efficiently.
In addition to constantly monitoring the data floor’s power and cooling needs, data center standards should approach every deployment with an eye toward efficiency and performance. The challenge is to deliver the optimal IT infrastructure setup for each customer without compromising performance elsewhere on the data floor. DCIM software, with its accumulated data on power and cooling usage, can help to ensure that every colocation customer is getting the most efficient deployment possible while also maintaining the overall health of the data center’s infrastructure.
Data centers necessarily use quite a lot of cable. Whether it’s bulky power cables or fiber-optic network cables, the facility must find ways to manage all that cabling effectively to make sure it all goes to the proper ports. While messy, unstructured cabling might be a viable solution for a very small on-premises data room in a private office, it’s completely unsuitable, and even dangerous, for even the smallest data centers. Cabling used in scalable infrastructure must be highly structured and organized if IT personnel are going to have any hope of managing it all.
Poorly organized cabling is not only messy and difficult to work with, but it can also create serious problems in a data center environment. Too many cables in a confined space can restrict air flow, putting more strain on both computing equipment and the facility’s cooling infrastructure. Inefficient cabling can also place unnecessary restrictions on deployments, which can make power distribution inefficiencies even worse.
Computer technology advances quickly. While the typical lifecycle of a server is about three to five years, more efficient designs that allow data centers to maximize their space and power usage can often make a piece of equipment obsolete before its lifecycle would otherwise suggest. With many data center standards pushing toward increased virtualization, there is a powerful incentive to replace older, less efficient servers.
But data centers don’t just need to think about cycling computing equipment. Power distribution units (PDUs), air handlers, and uninterruptible power supply (UPS) batteries all have an expected lifespan. Replacing these infrastructure elements on a regular schedule or controlled monitoring cycle allows facilities to maximize the efficiency of their data center operations and deliver superior performance to colocation customers.
By implementing a number of best practices, data centers can significantly improve their operations in terms of efficiency and performance. Colocation customers and MSP partners stand to benefit immensely from these practices, reaping the benefits of reduced energy costs and a more robust, reliable IT infrastructure.
Ross is a Regional Vice President, Operations at vXchnge and is responsible for managing all 14 data center locations. With more than 30 years of experience, Ross has managed data center construction, engineering, repair and maintenance, leading him to the emerging business of colocation.