GigaOm Radar for Edge Colocationv3.01

1. Summary

The arrival of 5G, the proliferation of internet of things (IoT) devices, and the increase in video streaming and augmented reality/virtual reality (AR/VR) are driving consumer expectations. For example, the opportunity for edge includes turning modern, on-premises workloads into SaaS delivery models, such as smart retail incorporating loss prevention, inventory tracking, automated checkout, and other features. As a result, businesses are looking for new ways to unlock revenue and increase brand loyalty by meeting the demand for a superior online quality of experience (QoE).

By moving data processing closer to the end user, edge colocation lowers jitter and latency and reduces traffic congestion, increasing the opportunity for dynamic personalization and providing a better QoE. Leveraging private networks (including 5G), software-defined networking (SDN), and network function virtualization, edge colocation data centers are generally smaller, decentralized facilities that provide compute and storage closer to where the data is generated and/or processed rather than backhauling it to regional colocation facilities or cloud data centers. In addition, edge colocation cross-connects often reduce latency and congestion by bypassing the public internet and tapping directly into fiber intersections and aggregation points, allowing for the setting up of complex service chains communicating across low-latency paths rather than having to “trombone” to a regional data center.

Deployed at standalone or collocated facilities, in a telco’s central offices, street cabinets, or cell sites, or attached to street furniture, edge colocation resources are offered by colocation service providers, managed service providers, and telecom service providers. As the market evolves and organizations look for edge processing with cloud economics, colocation providers are partnering to deliver a mix of capabilities designed to meet the needs of specific customer workloads. In addition, telecom service providers are increasingly partnering with public cloud providers to either resell edge cloud services to customers or consume them directly as part of their network infrastructure for fast, low-latency connectivity.

Representing features and capabilities widely adopted and well implemented in the industry, the following table stakes are the minimum features required for solutions to be included in the GigaOm Radar for edge colocation.

• Distributed edge locations: Edge colocation providers help businesses cope with the logistics nightmare of managing hardware installed in numerous distributed edge locations by renting out rack space near end users. Located in standalone facilities, on-premises at an enterprise, or in a telco’s central offices, street cabinets, or cell sites, the provider furnishes the structure, cooling, power, and physical security needed to maintain the equipment.
• Robust facility security: While many owned facilities have poor levels of security, with large numbers of employees allowed access and few checks in place to manage accessibility, edge colocation facilities are often located in remote areas and are generally unstaffed. A good colocation facility will have robust physical security in place, including video cameras, motion sensors, alarms, personnel check-ins, and physical isolation of tenants.
• Integrated facility monitoring: With edge colocation facilities generally unstaffed, remote monitoring of the physical status of the location—including power lines, temperature, humidity, and air quality—is essential. Therefore, each provider should implement IoT sensors and a building management system (BMS) to manage and monitor the facility’s equipment, including lighting, power, security, and heating, ventilation, and air-conditioning (HVAC) systems.
• Regional redundancy: Most edge colocation providers offer power backups, redundant networking, and other resources to increase the resiliency of workloads running in the facility. However, due to their distributed nature, individual edge colocation sites may be considered isolated failure domains. As a result, some providers offer lower redundancy of components at the site level. At a minimum, standby (load sharing) redundancy should be available at the regional level.

Once the table stakes are met, each provider is scored on key criteria and evaluation metrics. Key criteria are the basis on which organizations decide which provider(s) to partner with, while evaluation metrics determine the impact the provider’s capabilities may have on the organization.

Since, for the most part, colocation providers deploy and manage facilities in specific geographies, Figure 1 provides an overview of each provider covered in this report and the markets they service.

Figure 1. Colocation Providers and Geographical Coverage

This GigaOm Radar for Edge Colocation provides overviews of notable providers and their capabilities. The corresponding GigaOm report “Key Criteria for Evaluating Edge Colocation Solutions” outlines critical criteria and evaluation metrics for selecting an edge colocation partner. Together, these reports offer essential insights for optimizing workload processing, helping decision-makers evaluate provider capabilities before deciding where to invest.