Edge Data Center

---

An Edge Data Center is about putting computing and storage infrastructure out there in the world, close to the users, where it can be productive in real-time. This decentralized approach is comparable to Edge Computing.

Definitions

There are many definitions for an Edge Data Center. Any Data Center professional tends to define edge data center from the perspective of his own role, industry, or his own priorities.

1. The edge is where the internet of things (IoT), mobile computing and other time-sensitive applications will either thrive or languish. Data processing power at the edge of a network instead of in a cloud or a central data warehouse. Central clouds will continue to serve a purpose and will likely not be “blown away” by edge computing.
2. Edge is where the physical meets the digital. The request for the action is co-located with the computing, the analysis and the actual action delivered.
3. Monitoring at the edge, localized compute power for optimal performance that would help assist the local data processing, lower Latency and data transfer cost savings.

Basic Examples

Some of the basic examples of an Edge Data Center are:

1. The retailers try to close the gap between selling online and selling at brick and mortar locations there exist the Edge Data Center.
2. An Edge Data Center can include "Micro Data Center" deployed within small offices in the edge locations (not restricted to major cities), these data centers can be as small as something that fits under a desk.

Edge Computing vs Cloud Computing

Edge computing will not replace Cloud Computing, though the two approaches can complement each other. Using data from centralized computing and storage is called Cloud Computing. Cloud computing is a more general-purpose platform that can also work with purpose-built systems in the classical Client–Server model. Whereas computing at the Edge pushes almost all the data processing out to the Network EdgeNetworkEdge, close to the source of the data. Growth rates of both the methodologies can easily become unmanageable. So, the trade-off would be to divide the processing between the edge and the centralized system.

Key Characteristics of an Edge Data Center

In general, Edge Data Centers are:

1. Local: Data Centers are placed in tier-2 locations. This reduces Latency, Network Traffic, and costs by bringing data closer to the location they serve which improves uptime and availability. In remote locations, most Edge Data Centers are managed remotely.
2. Small: These data centers are smaller in size compared to their peer-systems. So, the Data Center Managers should utilize the space capacity efficiently.
3. Network Bandwidth: Edge Data Center typically has around 100 Gbps network backbone built with the most demanding user needs in mind.
4. Greater predictive scaling & lower Latency.
5. Built-in, not bolted-on, best-in-class network security.
6. Diverse and significant network federation capabilities.
7. Expert resources and time-tested solutions for managing complex networks.

Management Challenges

Depending on the deployment models, the Edge Data Center management challenges vary. Some common challenges which can be addressed with data-driven, remote visibility and control:

1. Making the most of Physical space and Power Capacity: When managing space remotely, color-coded floor map reports displaying cabinet weight, percentage full, budgeted and actual power, and other metrics can help you determine the best places to install assets at a glance and by using “what-if analysis” can help you understand the future impact of not only additions but also decommissions on your data center.
2. Directing technicians to complete changes properly: When you’re remotely managing multiple data centers, directing technicians on where to go and what to change can be one by using 3D floor map visualizations of each of your edge data centers to get a baseline of where your cabinets are situated. Device views allow you to show technicians where data and power ports are located on devices to ensure that installers make connections correctly.
3. Monitoring data center health across multiple locations: To address this challenge, we need both a holistic view. such as an enterprise dashboard showing real-time power and environmental health and events for all data centers on a single screen. These Metrics reduces the uptime.
4. Managing all assets and their connections across the entire data center deployment: Maintaining an accurate inventory of all your assets across these sites requires real-time views of cabinet contents, like servers, networking equipment, PDUs, and patch panels, as well as infrastructure devices and structured cabling.
5. Ensuring physical security: Restricting physical access to edge data centers can be difficult when managed remotely. Use reporting, audit logs, and surveillance feeds to monitor who has access to different areas and cabinets in your data center, how often these areas are accessed, and if the access attempts result in success or failure. This data can help you reduce risk by keeping track of who has access to your edge data centers and would be critical to forensic analysis following a breach.

Why Edge Data Center?

The need for edge data centers has grown markedly over the past five years. The edge computing market is expected to surge from $1.5 billion in 2017 to $6.7 billion in 2022, at a compound annual growth rate of 35%, according to a recent report by research firm “MarketsandMarkets”. Consumer demand for high bandwidth services and business demand for high capacity needs such as the Internet of Things (IoT) are just some of the primary drivers impacting demand for edge data centers. Other industries that see the benefit to edge data centers include retail, manufacturing, healthcare and financial firms.

Building an Edge Data Center

Building an Edge Data Center is very complex. There are a lot of things to consider and plan to ensure its success.

1. The very first thing to consider is selecting the perfect location to set up the Data Center. The data center equipment should be in an isolated environment with reliable heating, cooling, humidity control mechanisms and a fire-suppression system to ensure the safety of this expensive equipment. We would also need a security firewall for better protection.
2. For better reliability we would have to consider deploying backup power equipment, multiple connections to high-speed networks. Also, plan for the future, so it would be easy to migrate to high speeds up to 400 Gbps.

Micro Edge Data Centers (MEDC)

An extension to Edge Data Centers can be Micro Edge Data Centers. A Micro Edge Data Center would provide digital marketing services for a limited array of edge computing scenarios such as mobile logistics, 5G Software as a Service to support the specialized Internet of things. It would provide a physically and logically secured computing environment self-contained with all the storage, processing and networking required depending on the customers' applications. An MEDC can typically be deployed in less than six weeks and also reduce capital outlay, reduce the carbon footprint considerably, reduce energy consumption.

Practical Options for Today's Micro Data Centers

1. Compact sized advanced high strength steel fire resistant heavy gauge steel.
2. NFV, SDN, convergence, new chipset technology integration
3. Edge computing on-demand services use cases
4. Grey Box or White Box Offerings will be provided by multiple vendors of SD-WAN, NFV.
5. Virtual Network functions are desired.
6. Deep Packet Inspection for Cybersecurity and Network endpoint security a plus+.
7. NFV and SDN functional advances virtualized as software only should run only on white box servers and Virtualized enterprise vE-CPE (Customer Premises equipment) is an added plus.
8. Latency.
9. To either gain a competitive advantage or secure business: speed of deployment and testing.

Enabling Technologies for Emerging Data Centers

1. Resiliency and Edge Data Center Clustering : To support edge demand using multiple, micro-modular data centers. This involves usage of Load balancing, distribution, and failover at the IT, rather than power, infrastructure level.
2. Data Center Management as a Service (DMAAS): To monitor, measure and automatically manage the physical environment based on pre-set rules and thresholds.
3. Close-Coupled Cooling: Capable of dealing with wide temperature ranges.
4. Alternative Power Topologies: Using Nanogrids or Microgrids for on-site power generation, security and availability of a stable supply of power using renewable energy, fossil-fuel generator or fuel cells.
5. Microservers: Increased density, Disaggregate the CPU and Memory which allows for more efficient support, replacement and upgrade cycles of each independent microserver.

Common Problems

The common problems/issues in Edge environments can be, but not limited to Power, Cooling, Physical Security/Room Access, Ownership, Cabling.

Solutions to some Edge Problems

Any organization seeking or needing to identify and mitigate threats in their Edge Data Centers should seek out an integration with a thorough understanding of each of the following:

1. IT physical layers (servers, storage, & networking), with where and how they operate
2. Mechanical, Electrical Infrastructure
3. Cabling Infrastructure
4. Physical Security
5. Cyber Security

Corrected my article and addressed the issues highlighted by the previous reviewer.

This article "Edge Data Center" is from Wikipedia. The list of its authors can be seen in its historical and/or the page Edithistory:Edge Data Center. Articles copied from Draft Namespace on Wikipedia could be seen on the Draft Namespace of Wikipedia and not main one.

---