Python for Distributed Systems
summary
Builds a REST service that generates a Julia set image, measures real times and shows a C extension (called syon here) gives about a 4× speedup, then moves from a single Flask + Pillow process to a brokered architecture using shelf for persistence and zerum Q for messaging. Demonstrates the Global Interpreter Lock (Gill) problem with threads, implements a front-end/back-end via an in‑memory io module stream, and proves the brokered version is responsive and scalable.
AI-written summary · about
chapters (10)
- 0:00Python for distributed systems
- 1:07Global Interpreter Lock
- 18:25Local to scalable web service
- 27:54REST API with Flask for Julia set
- 41:37Scaling with parallel processing
- 53:20Async with threads and databases
- 1:09:16GIL impact on performance
- 1:13:49ZeroMQ and brokers
- 1:21:19Testing front-end and back-end
- 1:33:43Summary
related talks
30mBuilding a NoGIL Load Balancer in 30 minutes
47mThe many ways of parallel computing with Julia
24mEngage the Hyper-Python
32mJulia for Pythonistas
01h 10mTwisted up in a Distributed Tornado
42mBuilding a hosting platform with Python
42mBuilding a hosting platform with Python
26mPython Performance at Scale - Making Python Faster at Instagram