Machine learning enhances data centre operation

Data centers continue to increase in complexity as they are fragmented into edge computing, containerized deployments, hybrid IT and multi-cloud environments while still being interconnected to deliver applications. The level of sophistication needed to optimize these facilities and ensure application performance, requires operators to collect, harness and understand a tremendous amount of data from the facilities and IT stack. The IT industry needs intuitive tools to rapidly collect and analyze information, enabling data center operators to better understand how to manage workloads and their impact on critical facilities infrastructure.

“Regardless of the type of data center or business model, operators need to leverage analytics to minimize operating costs and understand the infrastructure where workloads are running,” said Enzo Greco, Chief Strategy Officer, Nlyte Software. “Nlyte has always been the thought leader that brought together facilities and IT operations.  The next step in this evolution is applying IBM Watson IoT’s leading machine-learning capabilities to head-off potential power and performance issues while also optimizing workload infrastructure operations and ultimately workload placement.”

Nlyte Machine Learning, powered by IBM Watson IoT, is addressing the data center analytics issue by collecting, normalizing and creating patterns of facilities and IT data and streaming the information to IBM Watson IoT. IBM Watson IoT then uses its machine-learning capabilities to extract predictive models and send the analysis back to Nlyte for a visual dashboard display of potential vulnerabilities, such as future hot server rows. With this information, data center administrators can proactively identify potential future issues and preemptively move server workloads. The net result is greater control of the infrastructure with more resiliency and increased reliability.