Every time you use your bank card to pay for a cappuccino, join a conference call from the road, or access a streaming service to kick back and relax, you’re benefitting from a big IT trend—the growth in edge computing.
Whether the goal is speeding the delivery of applications and content to remote users or collecting data from remote sites, edge sites are growing in number and importance. They range from regional and local data centers to the small computer rooms and communications centers in regional offices to network closets and Internet of Things (IoT) gateways in retail stores, school buildings, government offices and industrial sites.
And although they vary in size and complexity, they share some common traits. They are by their nature physically distant from the primary data center—and the team that supports it. Many are also in spaces that were never designed for IT, such as storerooms, factories, industrial sites and utility closets.
This situation can mean having to manage remote IT and network assets without visibility into that technology and relying on untrained on-site personnel for basic maintenance and troubleshooting. The result can be higher costs, lower reliability and increased downtime and stress on IT resources.
Fortunately, best practices have emerged that enable edge locations to be deployed quickly, managed remotely and maintained efficiently. Here are five to think about as you deploy or update edge sites.
1. Integrated Infrastructure
If you need to quickly deploy multiple edge sites or upgrade edge technology to support new applications, integrated infrastructure systems are an option that you should consider. Integrated solutions today range from complete pre-engineered and prefabricated modular data centers, such as the one deployed by T-Systems, to fully enclosed rows to individual racks of varying sizes from full racks to half and even quarter racks. These systems can be customized to site needs, comprise all of the infrastructure required (power, cooling, security and so on) to support IT equipment, and can be deployed much faster than integrating components on site allows.
Only the largest players will require a full data center at the edge, so most applications will be served by either a row or rack system. Today’s row-based systems, which can support up to 14 racks, not only include integrated power protection and monitoring but dedicated heat removal as well. Because heat is contained in the enclosed system, the cooling system operates at extremely high efficiency. Plus, an enclosed row can be deployed in an existing room, warehouse or industrial space with minimal modifications, saving time and money.
Smaller edge sites can take advantage of the range of single-rack systems that are available. These systems can include dedicated heat removal but don’t always require it if they are supporting a small number of IT assets. In this case, the UPS, power strips and remote-monitoring technology are configured to specifications and installed at the factory so the rack arrives on site IT ready.
An additional benefit of factory integrated row and rack systems is the physical security they provide. All equipment is contained in a lockable cabinet or enclosure. They can even be configured with multiple lockable bays in a single cabinet when third parties must share a rack—for example, in a retail store.
2. Enhanced Power Protection
Like in the data center, the reliability of edge locations depends on the power infrastructure. Unlike the data center, however, there has been a tendency in smaller edge locations to settle for “good enough” power protection.
With capacity and criticality growing, that situation is changing. Even small edge locations are moving to compact double-conversion UPS systems capable of filtering out the full range of power disturbances, such as those created by elevator motors or electrical storms, without draining batteries. These systems typically deliver higher reliability and require less-frequent battery replacement than the line-interactive systems they are replacing.
3. Remote Monitoring
Regardless of whether you’re using an integrated system or discrete components, remote visibility into edge sites is essential to monitoring performance and managing service. Communications cards embedded in the UPS can collect and communicate operating and environmental data to a central infrastructure-monitoring, infrastructure-management or building-management system. Continuous monitoring across multiple edge locations provides the visibility to better schedule preventive maintenance, react faster to events or outsource power system monitoring to a third-party service organization.
Intelligent power-distribution units enhance UPS monitoring by providing PDU-level and receptacle-level metering for insight into site and device power consumption. They also provide the ability to cycle equipment on and off through remote receptacle-level switching.
When paired with IT-management tools, such as serial consoles and KVMs that provide remote access and management of devices from a central location, these technologies offer visibility into and access to the power system as well as the equipment it supports.
4. Zero-Footprint Cooling
The 2016 Vertiv Cooling the Edge Survey identified two emerging trends related to small spaces: higher power densities owing to new equipment and growing criticality. Sites that once relied on building air conditioning to maintain environmental conditions are now outgrowing this strategy and require dedicated thermal management. More than 50 percent of the respondents to the survey said they would likely deploy dedicated cooling systems to network closets and server rooms in the year ahead.
Among the challenges they’ll face is floor-space limitations. Fortunately, zero-footprint thermal-management units are available that can manage heat removal without consuming floor space. They include compact and efficient ceiling or wall-mounted units that are configurable for the location.
When adding new equipment to edge locations or planning new locations, take the time to determine how much heat the equipment will generate and whether you’ve exceeded the threshold the building cooling system can handle. Call an expert before your system overheats.
5. Lifecycle Support
For most IT departments, it’s difficult to support a central location, let alone the multiple remote locations that characterize edge computing. Rather than strain existing personnel or hire additional staff, many organizations are outsourcing lifecycle support for their edge sites—from planning to installation and commissioning, remote monitoring and data analysis, preventive maintenance and battery replacement, to emergency response. Having a lifecycle support strategy in place to provide planned and emergency response can lighten the burden of maintaining a growing edge network.
Following these best practices can help you to gain control over network edge deployments with minimal strain, enabling the capacity, speed-of-deployment and reliability requirements of today’s distributed IT sites, all while keeping data centers and networks running reliably and efficiently. Factory integrated infrastructure, properly configured power and cooling, remote monitoring, and lifecycle support can all help take the complexity out of managing remote IT.
About the Author
Jack Pouchet is the vice president of market development at Vertiv. Based in Southern California, Jack works closely with major OEMs, large data center owners and operators, and leading mission-critical engineering firms to help improve day-to-day business and operational efficiencies while ensuring reliability, resiliency, and availability. Jack brings over twenty years of related OEM power-supply, power-generation, distribution and power-product sales and marketing experience to Vertiv, giving him a unique end-to-end perspective of the entire AC and DC power path.