NVIDIA System Management (NVSM) Grafana is a part of multinode software framework for monitoring NVIDIA DGX nodes cluster

Introduction

The NVSM Grafana Container is a Docker container that provides visualization capabilities for the NVSM multi-node architecture.
It works in conjunction with the Prometheus container to display system metrics and health information from the aggregator and compute nodes.

Key Features

• Real-time visualization of system metrics
• Customizable dashboards for cluster monitoring
• Integration with NVSM Prometheus data source
• Multi-node system health visualization
• Historical data analysis
• Alert visualization

Architecture

Core Components

1. Grafana Server

   • Web-based visualization platform
   • Dashboard management
   • User authentication
   • Data source integration

2. Pre-configured Dashboards

   • System health overview

3. Data Integration

   • Real-time data streaming
   • Historical data analysis

Container Specifications

System Requirements

• Part of the NVSM multi-node stack

Network Configuration

• Web Interface: 3000/tcp

Accessing Grafana

• Web Interface: http://[aggregator-host]:3000
• Default credentials: Configured during provisioning
• Dashboard access: Based on user permissions

Important Notes

• Part of the NVSM multi-node stack
• Should be deployed with other NVSM multinode containers
• Supports custom dashboard creation

References

• NVSM Documentation
• License
• Sources