Architect

Level: Architect | 15 Labs | Docker: python:3.11-slim

Deep-dive into CPython internals, advanced concurrency, security hardening, and production platform engineering. Each lab is independently runnable via Docker.

Quick Start

docker run -it --rm python:3.11-slim bash
pip install fastapi uvicorn pydantic pluggy cryptography prometheus-client opentelemetry-sdk

Lab Index

#	Lab	Topics	Key Verification
01	CPython Internals	PyObject, dis module, bytecode, frame objects, code objects, peephole optimizer	dis.dis() output on real function
02	Custom Import System	MetaPathFinder, PathEntryFinder, custom Loader, sys.meta_path, import hooks, pkgutil	Custom importer loading synthetic module
03	Descriptor Protocol	__get__/__set__/__delete__, data vs non-data, property internals, __set_name__, slots, validation	Typed descriptor with validation
04	Memory Allocator & GC	pymalloc arenas/pools/blocks, getsizeof, tracemalloc, gc module, weakref caches	tracemalloc top-5 allocators snapshot
05	C Extensions (ctypes)	CDLL, POINTER, Structure, Union, cast, libc qsort/malloc/free, CFUNCTYPE	ctypes calling libc qsort
06	Async Internals	Event loop, selectors, Future.await, Task scheduling, call_soon, async generators, contextvars	ContextVar request-scoped state demo
07	Advanced Metaclasses	__prepare__, metaclass new/init/call, conflict resolution, init_subclass, singleton, ORM registry	ORM-like metaclass field tracking
08	GIL-Free Concurrency	GIL impact, multiprocessing.Pool, ProcessPoolExecutor, shared memory, mmap IPC, as_completed	GIL vs multiprocessing speedup benchmark
09	JIT & Caching Patterns	lru_cache internals, cache_info/clear, wrapped, TTL decorator, disk cache, numpy caching	lru_cache hits/misses + TTL expiry
10	Protocol & Typing	typing.Protocol, runtime_checkable, Protocol.call, ParamSpec, TypeVarTuple, Concatenate, overload	mypy-compatible Protocol + isinstance
11	Distributed Task Queue	Redis-backed queue, task serialization, worker pool, exponential backoff, dead letter queue, result TTL	In-memory queue demo + Redis pattern
12	Observability Platform	OpenTelemetry TracerProvider, SpanKind, attributes, baggage; Prometheus Counter/Histogram/Gauge; structlog	Real span capture + Prometheus metrics
13	Plugin Framework	pluggy hookspec/hookimpl, entry points, init_subclass registry, versioning, DI, sandboxing	pluggy hook call chain
14	Security Hardening	secrets module, hmac.compare_digest, sha3_256/blake2b, cryptography Fernet + RSA + X25519, safe unpickling	Fernet encrypt/decrypt + RSA sign/verify
15	Capstone Platform	FastAPI + Pydantic v2 + import hook + descriptors + asyncio + tracemalloc + pluggy + Prometheus + Fernet + pytest	Full platform integration, 8+ tests

Available Packages (python:3.11-slim)

Package	Version
fastapi	0.135.1
numpy	2.4.2
pandas	3.0.1
pydantic	2.12.5
pytest	9.0.2
requests	2.32.5
uvicorn	0.41.0
rich	latest
anyio	latest
click	latest

Additional via pip install:

• pluggy — Plugin framework

• cryptography — Fernet, RSA, ECDH

• opentelemetry-sdk — Distributed tracing

• prometheus-client — Metrics

• structlog — Structured logging

• redis — Redis client (Lab 11)

Learning Path

Labs 01-05: CPython Deep Dive
  ├── 01: How Python runs your code
  ├── 02: Python's import machinery
  ├── 03: Attribute access protocol
  ├── 04: Memory management
  └── 05: C interop via ctypes

Labs 06-10: Advanced Python Patterns
  ├── 06: Async/await internals
  ├── 07: Metaclass engineering
  ├── 08: True parallelism
  ├── 09: Caching & memoization
  └── 10: Type system advanced

Labs 11-14: Production Infrastructure
  ├── 11: Distributed task queues
  ├── 12: Observability (traces/metrics/logs)
  ├── 13: Plugin architectures
  └── 14: Security hardening

Lab 15: Capstone
  └── Full production platform

Prerequisites

• Python Intermediate or Advanced track completed

• Familiarity with: classes, decorators, async/await, type hints

• Docker installed for isolation

Running Any Lab

# Interactive Docker session
docker run -it --rm python:3.11-slim bash

# Or run a specific snippet
docker run --rm python:3.11-slim bash -c "
  pip install PACKAGES -q 2>/dev/null
  python3 -c 'YOUR_CODE_HERE'
"