Building and testing resilient applications

on YouTube ↗28m

" summary

Shopify describes building a TCP proxy called Toxiproxy to inject latency and failures into tests and CI, discovers real bugs in Rails and drivers, and uses a resiliency matrix to convert many 500 cells into successful degradations. The team also implements fail‑fast tooling (Semian), plus circuit breakers and bulkheads, runs chaos experiments (Simian Army style), and reports that these concrete tools and tests materially reduce pages and improve survival during peak events.

AI-written summary · about

" chapters (10)

1. 0:00Start
2. 0:48Shopify
3. 3:37Resiliency
4. 5:43Fall Backs
5. 12:50Resiliency Matrix
6. 14:00Bugs in Rails
7. 20:09Circuit Breakers
8. 25:22The Simian Army
9. 25:47Resiliency Ladder of Maturity
10. 26:50Final Remarks

" related talks

• 26mResiliency Testing with ToxiproxyJake Pittis · LISA17
• 42mHow to stay alive even when others go down:...Florian Weingarten · ArrrrCamp 2015
• 16mHijacking Service Discovery to Simulate Dependency DegradationAbdulrahman Alhamali · SREcon25 Americas
• 25mResilience testing in modernization projectsŽivan Šinžar · HUSTEF 2023
• 40mResilience Testing - Failure is an OptionBartosz Czerwinski · TestCon Vilnius 2018
• 01h 02mPractical Chaos Engineering breaking things on purpose to make them more resilient against failureAdrian Hornsby · NDC Oslo 2019
• 42mPlanning To FailDave Gardner · PHP UK Conference 2013
• 21mTesting Rails at ScaleEmil Stolarsky · RailsConf 2016