Prod Lessons - Deployment Validation and Graceful Degradation

Key to Site Reliability Engineering is building frameworks and “guardrails” that enable the product to be developed safely. If patterns can be identified in outages and bugs, preventing those problems systematically gives SRE unparalleled leverage to improve stability. 

During this presentation, two frameworks that Pinterest has adopted from lessons learned in production will be discussed: deployment validation framework and product-informed graceful degradation. Both of these efforts have prevented hundreds of outages and saved many hours of engineering time.