The Everything Feed - All Packet Pushers Pods NAN101: Scaling Intel’s Data Centers with Network Automation (Sponsored)
Sep 24, 2025
In this discussion, Greg Botts, a senior network engineer at Intel with a rich background in networking, shares insights on automating the task of managing over 5,000 devices across 56 data centers. He talks about the transition from YAML files to a sophisticated rendering engine and the critical role of APIs in modernizing infrastructure. Greg also highlights Intel's choice of the open-source tool NetBox for better management and the exciting potential of AI in automating processes and enhancing network efficiency.
AI Snips
Chapters
Transcript
Episode notes
From YAML And DNS To A Homegrown Engine
- Intel began automation with a YAML file and DNS TXT records as a makeshift source of truth.
- A single engineer evolved that into a Python render engine and database before the team outgrew it.
The Unicorn That Built The Pipeline
- A single brilliant engineer built the initial Python 'render engine' that encoded provisioning logic and template rendering.
- That unicorn enabled years of progress but created dependence and risk when the engineer's role became central.
Shop For Supportable, Open, Low-Ops Tools
- When rebuilding automation, prioritize off-the-shelf, open standards, and low administrative burden.
- Choose vendor-agnostic tools and prefer projects with enterprise support options to reduce risk.