Network Performance and Monitoring Workshop

perfSONAR 5.1 is just around the corner. With this presentation, we want to update users and interested parties about the new features coming up with this version and the changes it will bring to your operation and usage of perfSONAR in the future.

TimeMap, a subtask from the GÉANT project WP6.T3 Monitoring, aims at developing solutions and methods for real-time analysis of latency and jitter measurements. Over the past two years, TimeMap has been applied as a monitoring service to the Géant backbone, but other NRENs and projects have also acquired it as a starting point for network latency real-time observability.

We will present the state of TimeMap, its architecture and the recent optimizations applied to make the tool more and more flexible and agile. We will introduce future developments to support new observability cases, such as T/F non-data services. Finally, we will discuss additional upcoming machine-learning methods that TimeMap adopts for abnormal event detection.

Presentation of TimeMap and perfSONAR usage by the GÉANT Operation Centre.

Wi-Fi is one of the most popular Internet access methods. Monitoring Wi-Fi networks to ensure their normal operation is vital. WiFiMon is a GÉANT service that offers open-source tools for efficient Wi-Fi network performance monitoring. The purpose of WiFiMon is to assist administrators in identifying underperforming segments within their networks.

Compared to previous WiFiMon presentations, we will describe newly introduced WiFiMon features, report our results from a recent WiFiMon pilot, outline our methods for supporting distributed configuration and control of hardware probes and briefly describe our recent efforts on performing anomaly detection on the collected measurements.

NetSage is an open-source, community-based measurement and monitoring tool that allows end users of any technical background the ability to access network performance data, using easy-to-understand, human-readable Grafana dashboards. These visualizations are open to anyone in the R&E community, enabling deeper insight into network utilization, traffic patterns, and end-to-end transfer performance across all science domains. In the last 18 months, we have made significant changes to the data collection process, front-end displays, as well as deployments, primarily focused on HPC centres in the USA. We have also developed a prototype to display Globus log data, giving previously unavailable insight into transfer patterns between Globus endpoints.

This talk will walk through the recent updates, demonstrate some of the new visualizations and data sources, discuss upcoming extensions for easier deployments, and present a roadmap for future development. Community feedback is welcomed and encouraged.