GPU Fabric ZTP & Commissioning

Zero-touch provisioning for GPU cluster networks: LLDP-based role assignment, lossless config push (PFC/ECN/DSCP) and a post-commissioning RDMA bandwidth benchmark.

Arista EOS

Cisco NX-OS

NVIDIA Cumulus

Open in Generator Open in Platform

You're viewing a preview

The full script, parameters, and execution console are available with a free account.

Already have access? Sign in →

Capabilities

LLDP topology discovery
Automatic spine/leaf/rail role assignment
Lossless config render (Jinja2)
Dry-run diff before push
NetBox commissioning sync
Post-push RDMA benchmark
Rollback on validation failure

Required inputs

Parameters the script accepts. Defaults shown; some are vendor- or context-gated.

Parameter	Type	Default	Notes
Rails per GPU node rails	number	8	—
Oversubscription oversubscription	string	1:1	—
Fabric MTU mtu	number	9216	—
Run RDMA benchmark benchmark	bool	true	—

Hint 1 — Name your vendors and OS versions

Mention the exact platforms you run (Arista EOS, Cisco NX-OS, NVIDIA Cumulus) so the generated ai/gpu dc logic uses the right CLI/NETCONF syntax.

Hint 2 — State your real thresholds

Provide concrete values for Rails per GPU node, Oversubscription, Fabric MTU instead of the defaults — they shape what counts as a fault.

Hint 3 — Prioritize the checks you need

This tool can lLDP topology discovery and automatic spine/leaf/rail role assignment. Ask for the subset relevant to your incident to keep output focused.

Hint 4 — Describe your inventory format

Tell the assistant whether your device list is CSV, YAML, or JSON and which columns it has, so parsing matches your data.

Sample inventory schema

Authoritative shape of the device/policy data this script consumes.

No bundled inventory sample. The script accepts standard device lists (CSV/YAML) with hostname, mgmt IP, vendor, and credentials.

Expected output

Reference terminal output the script should produce — used as a stylistic and structural target.

terminal

[2026-06-12] INFO  Discovering fabric via LLDP...
  Detected 4 spines, 16 leaves, 128 GPU links (8 rails/node)
[DRY-RUN] leaf03: +pfc priority 3 no-drop / +ecn 150 1500 / +mtu 9216 (12 lines)
[APPLY]   leaf03: config pushed, post-check OK
[BENCH]   gpu07↔gpu13: 392 Gb/s (98.0% of 400G line rate) OK
Summary: 16/16 leaves commissioned · 0 rollbacks · avg 97.4% line rate