Stage 2 of 7 · 8 lessons · ~7 weeks
Networking for engineers
Most production incidents that look mysterious are networking problems. Learn how IP, DNS, TCP, TLS and HTTP fit together and how to run a reverse proxy in front of an app.
Start: IP addresses, subnets and CIDRLessons· one lesson for each topic, in this order
- 1IP addresses, subnets and CIDRLesson
- 2TCP, UDP and portsLesson
- 3DNS records and resolutionLesson
- 4HTTP requests and responsesLesson
- 5TLS and certificatesLesson
- 6Firewalls and exposing only what you needLesson
- 7Nginx as reverse proxy and load balancerLesson
- 8Debugging with dig, curl, ss and tracerouteLesson
Start with
- VideoFreeComputer Networking: A Top-Down Approach (Video Lectures)
Jim Kurose and Keith Ross · Intermediate
Free lectures from the textbook authors; chapters 1 to 4 cover what DevOps needs
Practise
Keep handy
Also worth your time
- Load Balancer vs. Reverse Proxy vs. API Gateway: Demystifying Web Architectures (opens in a new tab)Read
Separates three components people often mix up, and what each one is for
Before moving on: Point a domain at a cloud VM, serve an app behind Nginx with a free TLS certificate, restrict the firewall to ports 22, 80 and 443, and debug a deliberately broken DNS record with dig.
Get good with dig, curl -v, ss and traceroute, because they answer most networking questions faster than any diagram. Subnetting and CIDR feel dry, but you will need them the first day you build a VPC.