Compute
AWS EC2 · GCP Compute Engine · Azure Virtual Machines — a rentable server, billed by the hour or second.
Lesson 2 of 4 · ⭐ Gap priority
Cloud Platforms & Core Ops
Every company names its infrastructure a little differently. Underneath, it's the same handful of primitives — learn those once instead of relearning them per employer.
The same primitives, different names
Four building blocks show up in almost every cloud-hosted system, whichever provider is underneath.
Object storage
AWS S3 · GCP Cloud Storage · Azure Blob Storage — flat, durable storage for files, addressed by key.
Serverless functions
AWS Lambda · GCP Cloud Functions · Azure Functions — run-on-trigger code that scales to zero.
Managed Kubernetes
AWS EKS · GCP GKE · Azure AKS — a control plane the provider runs for you; you still manage the workloads.
CI/CD pipeline stages
Jenkins, GitHub Actions, and CircleCI all implement the same four stages — the tool changes, the shape doesn't.
Build
Compile or package the app into a deployable artifact — a container image, a binary, a static bundle.
Test
Run automated checks (unit, integration, lint) against the artifact before it goes anywhere near production.
Deploy
Ship the artifact to an environment, often via a rolling or blue/green strategy to avoid downtime.
Rollback
Revert to the last known-good artifact fast, without re-running the whole pipeline, when a deploy misbehaves.
Linux & networking fundamentals
Every service you'll touch is a process listening on a port, reachable because DNS and routing agree on how to get there.
DNS turns a name into an IP address; a resolvercaches that answer for the record's TTL, which is exactly why a change can take a while to show up everywhere. Once you have an IP, the OS decides how to route to it, and the remote process is either listening on the port you're hitting or it isn't. Four commands cover most first-pass diagnosis: dig (what does this name resolve to, and where), curl -v (what actually happens on the wire), ss -tlnp (what's listening locally), and ping/traceroute (basic reachability).
Exercises
Look at the sanitized dig/curl output below, then work through both exercises.
$ dig +short api.internal.example.com
10.0.4.12
$ curl -v https://api.internal.example.com/healthz
* Trying 10.0.4.12:443...
* connect to 10.0.4.12 port 443 failed: Connection refused
* Failed to connect to api.internal.example.com port 443
$ dig +short api.internal.example.com @8.8.8.8
10.0.4.55Concept check
Answer in your head first, then click a question to reveal the answer.
Scenario checklist
Work through the sample dig/curl output and check each item off. Progress is saved in your browser.
Go deeper
These primers are a starting point. Read the primary sources to build real fluency.
Google Cloud: AWS to Azure to GCP service comparisonA living table mapping equivalent services across the three major cloudsRed Hat: What is CI/CD?The pipeline stages and why each one exists, vendor-neutralCloudflare Learning: What is DNS?Resolution, caching, and TTLs — the part that bites you at the worst timeGitHub Docs: Understanding GitHub ActionsOne concrete implementation of the build/test/deploy stages above
Training Notes
What building this project taught me.
This is a flagged gap-priority module
Cloud infrastructure fundamentals was Tom's #3 self-identified gap across a scan of 30+ job descriptions — sequenced second in this trail, right after the top-priority containers module.
This module stays at the fundamentals layer
Deep dives on any single provider's console or IAM model are out of scope here. The goal is enough shared vocabulary to onboard onto whichever provider a given company uses, fast — not certification-level depth in one of them.