Top 10 Fails on a Real Highload Project
summary
A review of the top 10 architecture failures found on a real high-load project, describing concrete systemic mistakes and their consequences. Covers real examples and lessons learned for building scalable systems, aimed at improving architecture, fault handling and operational practices in high-load environments.
AI-written summary · about
chapters (12)
- 0:00Highload systems intro
- 1:27User ID choice failure
- 7:25Zero-downtime migration issues
- 11:07Table cleanup under load
- 15:24Consul cluster release impact
- 16:28Cache invalidation and hashing issues
- 18:36Redis GC losses
- 21:25Layout errors and image resizing
- 22:51Apache hang from socket server
- 23:552M emails lost from bad data
- 25:00Stats collection problems
- 28:54Highload systems conclusions
related talks
47mTop infrastructure mistakes that hinder high loads
43mFunctional Testing for High-Load Projects
16mClosing remarks
01h 42mRoasting Architecture Case Studies
44mWhat Is a Highload System? Are All Highload Systems Similar and Do They Share the Same Problems?
18mDemonstrating your Architecture
50mCracks in the Foundation: Operating Cloud Foundry...
41m12 Factors of Pain and Suffering