NVIDIA Mission Control DGX SuperPOD Ethernet North-South Network Configuration Guide

Contents:

• Introduction
  • Reference Documentation
  • Downloads
  • Physical Hardware
• Networking Planning and Design
  • Logical Network Design
  • Ethernet Networking (North-to-South)
  • Subnet Breakout per POD
  • Subnetting by DGX Model
  • BGP ASN 4-Byte Range
  • Ethernet Planning
    • Confirm Accurate Hardware Models and Quantities
    • Confirm You Have the Correct Cumulus Linux Software Version
    • Obtain Customer Network Requirements
  • DGX Compute VLAN
  • OOB VLAN / VNI
  • Edge Network Handoff
    • InBand and OOB Handoff Separation
• How to Format Point-to-Point (P2P)
• Understanding the Application Workflow
  • Workflow Diagram
• BCM Bootstrap Workflow
  • Bootstrap Process Overview
  • Bootstrap Stages
  • Troubleshooting
  • Completion Criteria
• Configure Cluster Ethernet Networking
  • Deployment Process Options
  • Prerequisite Stage
  • Pre-Run Process
  • Manual Process
    • Configuring the Cumulus Switch using the Template
  • Semi-Automated Process (DHCP/ZTP)
  • Fully Automated Process (BCM/ZTP)
    • On-Site Process
  • Alternative Deployment Options
    • Option 1: NO bcm-netautogen, YES bcm-pod-install
    • Option 2: NO bcm-netautogen, NO bcm-pod-install
• Validate Ethernet Cluster Networking
  • Platform Health
  • Cabling Quality (L1)
  • Fabric Underlay Routing
    • Fabric Overlay Routing (EVPN)
    • Show Ethernet Segment Information
    • Customer Network Reachability
    • Negative Testing
• NVLink Switch Planning, Design, and Deployment Guide using Base Command Manager (BCM)
  • NVLink Switch Planning
    • Hardware Components and Models
    • Network and Connectivity Requirements
  • NVLink Switch Design
    • Physical Connectivity
    • Network Allocation Diagram
  • NVLink Switch Integration with BCM Verification
    • Verify the NVLink Switch Configuration
    • Updating the NVLink Switch to the Latest Firmware
• Glossary of Terms
• Appendix
  • Section 1.1: siteinfo.yaml
  • Section 1.2: Standard Point-to-Point (P2P) Column Header
  • Section 1.3: Standard Worksheet Naming
  • Section 1.4: Connection Type
  • Section 1.5: Standard Naming Conventions for Network Components
  • Section 1.6: Example Point-to-Point (P2P) format
  • Section 2.1: GB200 and GB300 Rack Inventory
  • How to pull rack inventory from the Splunk database