Watch a 26-minute conference talk from the INSAIT 2022 Conference where Dr. Rupak Majumdar from Max Planck Institute and AWS explores the effectiveness of random testing in complex distributed programs. Learn how combinatorial insights can prove properties of random testing procedures, understand the relationship between testing coverage and the probabilistic method, and discover why fuzzing works well as a guided random testing approach. Delve into practical examples involving distributed systems with partition faults, k-splitting families, and Jepsen testing tool applications. Gain valuable insights into computational challenges associated with random testing methodologies and their implications for ensuring correct distributed systems.
How to Randomly Test Programs and Why Random Tests Are Good
Overview
Syllabus
Intro
How can we ensure Correct Distributed Systems?
Tool: Jepsen
Tests and Coverage
Probabilistic Method
Ninjas in Training
From Training Ninjas to Distributed Systems with Partition Faults
Small k-Splitting Families
Partitions and Jepsen
Fuzzing: Guided Random Testing
Why does Fuzzing work well?
An Example (from the last talk)
Summary
Taught by
INSAIT Institute
