Designing a scalable and distributed application
summary
A runnable geo-distributed page hit counter is assembled with nginx and uWSGI frontends, local queues collocated with collectors, and consul for service discovery and a replicated key/value store. Processors pull jobs from local queues, increment local counters, and Consulagents sync counters across regions so the collectors can sum them. A live demo shows automatic registration, health checks, immediate config propagation and that the system survives a simulated inter‑datacenter partition while keeping counts consistent.
AI-written summary · about
chapters (10)
- 0:00Start
- 0:39Designing a scalable and distributed application EuroPython 2015
- 1:46Geo distributed page hit counter web application
- 3:24Automatic resizing no manual application re configuration
- 6:37Isolated components
- 7:23Microservices
- 8:22Communication reliability network latency and failure
- 11:09processor consumes jobs increments a counter
- 15:52Service discovery automated scaling and fault toleran
- 18:48UWSGI consul plugin seamless service registration / health check
related talks
47mBuilding Scalable Web Apps
42mBuilding a hosting platform with Python
42mBuilding a hosting platform with Python
15mMonoceros: Faster and Predictable Services through In-pod...
19mOptimizing Data Retrieval with Python Celery
36mAn Introduction to Distributed Task Queue
24mDistributed Web Scraping in Python
01h 34mPython for Distributed Systems