peshkova - Fotolia
- Jean DerGurahian, Features and E-Zine Editor
In the aircraft industry, engineers can model the wing of a plane; evaluate how airflow, the weight of the wing material and the fuel carried in the wing all affect the plane's performance; then determine whether the wing design meets strength requirements. All of this is accomplished using computing tools and analytics. Engineers assess the final results. They're not sitting at desktops drawing wing models or using calculators to gauge airflow.
The networking industry, while not quite at that level of modeling and analyzing, is headed that way, as enterprises increasingly want to derive meaningful and actionable information from network data analytics.
Technologies and concepts including software-defined networking, software-defined WAN and cloud computing are affecting network design and how companies do business. Networks, like airplanes, have a lot of complicated, moving parts, and we are just at the beginning of using new network data analytics tools that incorporate machine learning and artificial intelligence to help study every aspect of the network. What we know now is that we have a lot more to learn.
Technologies like network orchestration and intent-based networking are pushing network managers toward the use of broader and deeper analytics tools, according to Andre Kindness, a principal analyst at Forrester Research. The more complex and automated networks become, the more network managers will need to use analytics to keep track of how those elements are performing. It will be done in the name of efficiency, Kindness said.
Right now, networking is still reactive to issue, he added. The goal for analytics is to help make network managers more proactive.
At some point, engineers will be able to create a network model much like the airplane wing model. On the human side, reaching that point will require more communication and collaboration among enterprise network managers, the entire IT department, end users, vendors and software developers.
Network analytics is only as good as the data that feeds it, and data must come from all sources if it's going to create a comprehensive, holistic picture of the network. That means bringing together data from disparate infrastructure, application, workflow, device and product sources that in the past have been left in siloes.
On top of that, network engineers have to be as high-performing and reliable as the networks they run. This means broadening their understanding of software-defined networking, machine learning, AI and software development. The good news about network data analytics is that it helps do the job faster. But nothing can replace the human element in understanding why getting the job done faster and better is important.
Isaac Asimov said it best in his short story series I, Robot: "The machine is only a tool after all, which can help humanity progress faster by taking some of the burdens of calculations and interpretations off its back. The task of the human brain remains what it has always been; that of discovering new data to be analyzed, and of devising new concepts to be tested."
Dig Deeper on Network management software and network analytics
An overview of Apache Airflow for workflow management
Cloudera adds data engineering capability to enable DataOps
Coronavirus: Using computer simulation to minimise cross-infection
Nokia inks TIM Brazil IoT contract and supplies Deutsche Bahn 5G automated rail operation