Provisioning the fabric is not enough; visibility into the solution is key. With Augtera, Dell Enterprise SONiC integrates monitoring and deep insight into the AI solution.
The integration uses open-standards traffic monitoring technology such as sFlow, everFlow, SNMP tracking, and SONiC gRPC, gNMI Telemetry.
Augtera SONiC AI is a multi-vendor and multi-technology solution that provides monitoring and anomaly detection of Enterprise SONiC Distribution networks in a greenfield or brownfield environment using standards-based configuration and monitoring techniques. SONiC AI uses machine learning (ML) on networking data and helps keep the focus on the business while minimizing expenses.
To tackle SONiC network operations challenges, Augtera Networks tapped into AI and ML. The effort to incorporate AI and network management resulted in the industry's first AI platform for Enterprise SONiC networks called SONiC AI. SONiC AI provides a holistic solution in addressing network operations.
Augtera SONiC AI brings the benefits of AI-augmented operations to physical, virtual, and cloud network environments and achieves the following outcomes:- Integration and validation with Enterprise SONiC Distribution by Dell Technologies
- As data center networking moves towards Open Networking with the implementation of open-source NOS, Augtera works with Enterprise SONiC to make it easy to configure and manage SONiC using SONiC AI. Some key integrations include:
- Integration with Enterprise SONiC gRPC/gNMI telemetry models for comprehensive monitoring and AI-based anomaly detection
- Integration and advanced classification of Enterprise SONiC syslog to generate alarms
- Integration with Enterprise SONiC Everflow for on-demand flow analysis
- Advanced Machine Learning applications
-
ML is suited to automatically recognize patterns from data with minimal human intervention. With network devices supporting telemetry streaming, large amounts of data are produced and collected, but the backend systems are not able to analyze this amount of data.
The collected data leaves operators searching for needles in a large data haystack. However, network infrastructure is a specialized and complex source of data that makes it hard to apply ML to this data while deriving operational value. - Comprehensive converged monitoring and automation solution
-
SONiC AI can visualize the topology, metric charts, tables, and graphs of the devices, logs, and events, enabling SONiC AI to generate alerts and notifications in the network.
Thus, it can substitute the functions of a traditional network management system (NMS) while providing the key differentiation of identifying highly complex correlated anomalies using AI/ML. This capability makes SONiC AI the "single pane of glass" for visibility of the entire network. Augtera pulls data not only from the network devices, but also from other tools in the environment, which helps in the consolidation of data, and prevents further tool sprawl.