One of the challenges facing multi-access edge computing (MEC) is data management. MEC is a type of distributed cloud established from local sources that situates computing closer to end users than the cloud. MEC delivers a reliable connection, offers low latency, and can handle a substantial magnitude of computing sources. Establishing computing closer to the end-user improves computing in two fundamental ways: it reduces network congestion, and it boosts application performance.
The ability to localize computing meets users’ networking demands as they connect additional devices to the internet. Users’ auxiliary devices, such as those used for creating smart homes, smart cities, and other Internet of Thing (IoT) devices, will generate and transmit a staggering amount of data. According to IDC, the data “is doubling in size every two years, and by 2020 the digital universe — the data we create and copy annually — will reach 44 zettabytes, or 44 trillion gigabytes.” This data increase means that current performance management tools must evolve to sort, filter, and analyze that amount of data.
The tools that gather and interpret data to pinpoint any issues within a network’s or application’s performance is known as performance management. Performance management includes three primary tools to conduct the gathering and root-cause analysis: application performance management (APM), network performance management (NPM), and unified performance management (UPM). As computing advances, UPM is emerging as the leading tool for monitoring both applications and networks because it is a converged approach to APM and NPM. UPM presents network administrators with a more comprehensive view of their services using a single pane of glass view versus them checking multiple APM and NPM dashboards for the same insight.
Additionally, monitoring edge computing is more painless than monitoring the cloud. As ETSI mentions, “due to the virtualized characteristics of MEC, it has never been easier to monitor performance and resource needs of an application.” However, even with MEC improving performance management processes and the value of the converged performance management solution, the amount of data that connected devices send to the edge presents a unique challenge to performance management tools and IT personnel.
MEC Performance Management: How the Tools Should Adapt
The tools of MEC Performance management need to operate in the following ways:
- Provide end-to-end visibility across the edge computing network;
- Offer real-time visibility;
- “Seamlessly monitor packet information in both current and historical data, including all the traffic from Apps, network protocols, VPN-to-edge,” according to Lanner;
- Proactively troubleshoot;
- And implement machine learning (ML) for automation and predictive analytics. ML will learn from past data that signaled network issues to predict future problems before they occur.