Digital Transformation through GitOps and Infrastructure as Code
summary
Converted running configs into machine-readable YAML/JSON and Jinja templates stored in GitHub Enterprise, automated deployments via Python/Ansible and device APIs (with fallbacks to SSH and tools like Netmiko/NAPALM), and run CI/CD tests and QA so a team of ~8 engineers manages ~3,500 routers, 600 firewalls and ~40 data centers for over 6.5 million customers. The pipeline supports Jinja templating, remote zero-touch builds (e.g., 6 routers, 20 switches, 10–12 firewalls per site), faster provisioning, fewer human errors, and full versioned auditability.
AI-written summary · about
chapters (12)
- 0:00Introduction to network programmability
- 4:34Evolution of network engineering
- 10:15Challenges of manual network operations
- 16:29Benefits of network automation
- 22:12GitOps and source of truth
- 25:43Desired state and troubleshooting
- 31:18Version controlled infrastructure
- 36:20Quick infrastructure provisioning
- 41:15Network configurations as code
- 43:45Automation pipeline walkthrough
- 47:18GitOps beyond infrastructure
- 51:51Q&A and resources
related talks
27mFrom Spreadsheets to "Everything as Code"
55mGitOps, Fostering Next-Generation Operations Teams
30mDevOps: Supporting aircraft from hanger to take-off
29mWe Have Always Done It This Way! Now Let’s Try Something Completely Different
38mAutomating and Managing an IoT Fleet Using Git
29mThe Evolution of GitOps in Platform Engineering
15mGitOps in the Real World: Opportunities for Developer Experience Improvement
43mHow GitOps Enabled Us to Answer an Unanswerable Question