Skip to content

DevOps — Overview

This section is a practical tour of modern DevOps: the culture and measurements behind it, the operating-system and network fundamentals everything rests on, and the tooling for building, shipping, running, observing, and securing software. Lectures contain real configuration (bash, YAML, HCL, Dockerfiles, PromQL) rather than only theory. The ecosystem moves quickly; lectures state what is stable (concepts) versus fast-moving (product names, versions, licensing).

  1. DevOps Culture & Lifecycle: DevOps vs SRE vs platform engineering, DORA metrics, CALMS, value streams, shift-left.
  2. Linux & Networking for DevOps: processes, permissions, systemd, signals, TCP/IP, DNS, TLS, and a debugging toolbox.
  3. Shell Scripting & Automation: safe bash, pipes, jq, cron and systemd timers, Makefiles, and when to switch to Python.
  4. Git Deep Dive & Branching Strategies: objects and refs, merge vs rebase, trunk-based development, hooks, signing, recovery.
  5. CI/CD Pipelines: pipeline principles, GitHub Actions in depth, OIDC to the cloud, deployment strategies, feature flags.
  6. Containers & Docker: namespaces and cgroups, images and layers, Dockerfile best practices, Compose, image security.
  7. Kubernetes Fundamentals: architecture, core objects, probes, resources, scheduling, storage, RBAC, debugging.
  8. Kubernetes in Production: Helm, Kustomize, autoscaling, operators, service mesh, CNI, GitOps, upgrades, cost.
  9. Infrastructure as Code: Terraform/OpenTofu state and modules, Pulumi, Ansible, immutable infrastructure, policy as code.
  10. Cloud Fundamentals: service models, IAM, VPC networking, primitives across AWS/GCP/Azure, FinOps, multi-region.
  11. Observability & Monitoring: metrics, logs, traces, Prometheus and Grafana, OpenTelemetry, alerting, SLOs.
  12. Incident Response & Reliability Engineering: on-call, runbooks, postmortems, chaos engineering, capacity planning, DR.
  13. DevSecOps & Supply-Chain Security: secrets, SAST/DAST/SCA, SBOMs, Sigstore, SLSA, least privilege, hardening.
  14. MLOps & LLMOps: experiment tracking, versioning, feature stores, serving, drift, GPUs on Kubernetes, LLM gateways and evals.
  15. Deploying Static Sites on Vercel: how Vercel builds Astro, previews, env vars, domains, caching, alternatives.
  • Comfortable with the command line and one programming language (Python or similar).
  • Basic networking vocabulary (IP address, port, HTTP) is helpful; lecture 2 fills gaps.
  • For the MLOps lecture, a passing familiarity with training and serving a model; see AI Foundations and Inference & Serving.
  • A laptop with Docker (or Podman), Git, and optionally kind and a free cloud-account sandbox for the hands-on exercises.

Roughly 40-55 hours including exercises; the order below builds each layer on the previous one.

Week Lectures Focus Hands-on goal
1 1-3 Culture, Linux, networking, scripting Write a hardened systemd service and a safe backup script
2 4-5 Git and CI/CD A PR pipeline with caching, matrix, and OIDC deploy
3 6-7 Containers and Kubernetes basics Containerize an app, deploy it on a kind cluster with probes
4 8-9 Production Kubernetes and IaC Argo CD or Flux deploying from Git; Terraform a small cloud stack
5 10-11 Cloud and observability VPC plus bucket via IaC; Prometheus, Grafana, an SLO and burn-rate alert
6 12-13 Reliability and security Run a game day; add scanning, SBOM, and signing to the pipeline
7 14-15 MLOps and shipping a site An eval gate in CI; deploy this wiki with a custom domain

Shorter path (about 20 hours): lectures 1, 2, 4, 5, 6, 7, 11 give a working DevOps foundation. Read 8 and 9 once you are running real workloads, 12 before you carry a pager, and 13 before anything handles production data.