Issue link: https://www.e-digitaleditions.com/i/1414667
October/November/December 2021 I 23 However, IoT is not the only 5G game in town. 5G will primarily be an enterprise-driven business case, characterized by operators clamoring to form partnerships across and within different industrial sectors, including healthcare, automotive, agriculture, construction, media and entertainment, finance, mining, and many others. 5G networking will be mission critical in many of these industrial use cases. Due to both geographic and application performance necessity, a multi-data center application environment should be considered. VISIBILITY IS NEEDED TO MANAGE COMPLEXITY The old operation's truism that states "what is not measured cannot be managed" holds true for today's highly complex, dynamic, massively scalable, and highly agile data center reality. Visibility (i.e., measurement) is gained through both service and network monitoring. In fact, both are needed to get a true picture of the customer's quality of experience (QoE) and the issues that negatively impact it. It is easy to think that managing DCI is as simple as occasionally monitoring some or all of the optical fibers with an optical time-domain reflectometer (OTDR). After all, this is often offered as simply a dark fiber service. However, this may not tell the complete story. For example, the case of a cell tower backhaul link configured to operate at 10Gb/s (CPRI-8) suddenly fell back to 622 Mb/s (CPRI-1). The mobile operator detected the issue as traffic congestion, but after a lengthy investi- gation determined the problem to be a microbend in a patch cord caused by maintenance activity. In this case, being able to correlate higher layer network and service monitoring with optical link monitoring would have identified the route cause immediately, saving both time and customer frustration. This is a relatively simple example, but in a multi-data center network, with literally hundreds of thousands of fiber connect- ions and millions of miles of fiber cable spanning tens or even hundreds of data centers, this quickly becomes an untenable problem that warrants more advanced performance monitoring solutions. Seeking cost savings and an economic advantage, operators such as Rakuten, Verizon, DISH, and Three UK are leveraging third-party cloud providers to deploy some or all of their core and mobile edge network functions. 5G opens even more opportunities for traditional oper- ators to do the same, as well as for non-telecom entities to provide 5G-enabled telecom services. All of this adds up to an increased reliance on data centers and DCI to support the mission-critical services 5G promises. Since there is no reward without some risk, DCI providers are increasingly required to offer highly stringent service level agreements (SLAs) to win the web- scale and service provider interconnect business. These SLAs come with significant financial penalties for failing to meet the contracted requirements, but perhaps even more worrisome is the damage to brand reputation if it is determined the DCI services were at fault for a major network outage. DCI providers are increasingly required to offer highly stringent service level agreements (SLAs) to win the webscale and service provider interconnect business. Having complete visibility into the performance of the services provided and any possible negative impact on QoE is no longer an option. For DCI pro- viders, this must become a business imperative! This is where the benefits of AI-driven, multi-layered service assurance really stand out—automatically identifying or mitigating SLA breaches that help to deliver a better QoE for everyone.