14 parts · 43 chapters
Node.js Internals
You already know JavaScript. This course is about the runtime: the C++ and C underneath your code, how a request moves through V8, libuv and the kernel, how Node fails under load, and how to run it at staff level. Every claim comes with something to run: a flag, a script, a profile.
Fourteen parts, following the syllabus outline. What Node actually is; V8 in depth; libuv and the event loop for real; streams, buffers and backpressure; modules and bootstrap; the native layer; concurrency and multiple cores; networking and API design; databases from Node; performance engineering; running Node in production with what it is paired with at scale; observability; tooling, testing and security; and a capstone with an interview bank.
V8Ignition, Sparkplug, Maglev, TurboFan, hidden classes, inline caches, deopts and the garbage collector.
libuvThe seven phases, nextTick and microtasks, the threadpool, and event loop utilisation.
streamsBuffers, backpressure, pipeline, Web Streams and custom transforms.
modulesCommonJS, ESM, loaders, the dual package hazard, single executable apps.
nativeBindings, N-API, addons, WASM and FFI.
concurrencyWorkers, Atomics, cluster, child processes and when to use which.
networkingllhttp, HTTP/1 to 3, undici, TLS, WebSockets, frameworks, graceful shutdown.
productionContainers, memory limits, serverless, the production stack, OpenTelemetry, security.
00
What Node Actually Is
The 2009 origin and the C10K problem · The architecture: V8, libuv and the C libraries · Release lines, governance, and Node vs Deno vs Bun
3 ch · ~18 min01V8, In Depth
The compilation pipeline and deoptimisation · Hidden classes, inline caches and megamorphism · Garbage collection, heap layout and finding leaks
3 ch · ~18 min02libuv and the Event Loop, For Real
Loops, handles, requests and the phases · nextTick, microtasks and phase queues · Platform IO and the threadpool · Blocking the loop, ELU and timers
4 ch · ~24 min03Streams, Buffers and Backpressure
Buffers, typed arrays and zero-copy · Stream types, state machines and backpressure · pipeline, Web Streams, async iterators and performance
3 ch · ~18 min04Modules, Loading and Bootstrap
Bootstrap and CommonJS · ESM, the dual package hazard and exports maps · Loaders, startup performance and single executables
3 ch · ~18 min05The Native Layer and Extending Node
The binding layer · Node-API and a native addon end to end · WASM, FFI and embedding
3 ch · ~18 min06Concurrency and Multi-Core
Workers, messages and shared memory · cluster, child processes and the decision matrix
2 ch · ~12 min07Networking and API Design in Node
The http stack, keep-alive, HTTP/2 and HTTP/3 · undici, TLS, WebSockets and SSE · API design, frameworks, validation and resilience · Graceful shutdown done correctly
4 ch · ~24 min08Databases from Node
Drivers and the wire protocol · Pooling, transactions and async context · Query optimisation and caching from the app side
3 ch · ~18 min09Performance Engineering
Benchmarking and CPU profiling · Memory, leaks and production diagnostics · Optimisation techniques, and when to go native
3 ch · ~18 min10Running Node in Production
Containers, signals and memory limits · Topologies, serverless, edge and zero-downtime deploys · The production stack around Node
3 ch · ~18 min11Observability
Logging and metrics · OpenTelemetry and tracing across async boundaries · Errors, unhandled rejections and continuous profiling
3 ch · ~18 min12Code Quality, Tooling, Testing and Security
Toolchain, TypeScript and linting · Testing strategy, real databases and fuzzing · Monorepos, dependencies and Node security
3 ch · ~18 min13Build and Expert Mode
The capstone and reading the source · Sixty questions, in six groups · War stories
3 ch · ~18 minBuilt on JavaScript, Computers and ConcurrencyThis course assumes JavaScript the language (the JS course) and uses the machine model from How Computers Work. The Databases parts lean on MySQL, SQL and PostgreSQL Internals.