Augtera Networks Proof of Concept: Proactive Detection and Isolation of Congestion in a Dell NVMe IP-Based SAN using Augtera Networks Machine Learning
Dell existing storage architectures and next-generation architecture with NVMe/TCP solution enable architects to build highly scalable storage environments that allow large-scale deployments and operations over distances.
Benefits of using regular TCP/IP based networks are that it is highly scalable, supports disaggregated infrastructure, and reduces cost.
As organizations seek to take advantage of these benefits it becomes critical to make sure the underlying IP infrastructure is performing as expected. Specifically, that the underlay has no congestion or errors.
Augtera Networks will show in the ONUG POC how Augtera Network AI platform provides a simple workflow to enable this. This POC will demonstrate how:
Proactive detection and notification of congestion can be achieved using native sFlow telemetry from switches coupled with machine learning capabilities from Augtera.
Augtera analytics and visualizations can be used to determine the TCP flows that are impacted by the congestion
Operators can gain an end-to-end view including the servers and the Data Center fabric to isolate whether the congestion is caused by the fabric, the servers or the interconnect between the servers and the fabric.
Operators can be notified of congestions events using collaboration tools such as Slack or enterprise ticketing tools such as ServiceNow
Related Links:
Dell NVME/TCP Overview: https://infohub.delltechnologies.com/l/nvme-nvme-tcp-and-dell-smartfabric-storage-software-overview-ip-san-solution-primer-1/overview-2551
Augtera Data Sheet: https://augtera.com/wp-content/uploads/2022/03/Augtera_Get-Ahead-of-the-Incident-with-Network-AI_r3c.pdf [augtera.com]
Augtera Data Center Solution Brief: https://augtera.com/wp-content/uploads/2022/07/Augtera-Data-Center-Solution-Brief.pdf [augtera.com]
