Application-Level Chaos Engineering in JVM
summary
A research toolset instruments JVM bytecode with a Java agent to inject exceptions and measure real production behavior: ChaosMachine throws exceptions at the start of try blocks, classifies catch blocks as resilient / silent / debug and finds fragile, sensitive, and immunized perturbation points using TTorrent. A companion system, TripleAgent, combines monitoring, perturbation and failure‑oblivious interception at the method level to suggest automated resilience fixes with modest CPU overhead. A Docker pipeline augments base images for easy deployment; code is on github.com/KTH/royal-chaos.
AI-written summary · about
chapters (17)
- 0:00Preamble - The Space of Chaos Engineering
- 1:37Royal-Chaos @ Github
- 2:03Chaos Machine
- 5:02The Overview of ChaosMachine
- 6:15ChaosMachine - Hypotheses
- 8:26What Can be Learned
- 9:02Experiments on TTorrent
- 10:10TripleAgent
- 11:33A Chinese Kungfu in Chaos Engineering - httops://en.wikipedia.org/wiki/Zhoun_Butong
- 12:39TripleAgent - Example
- 15:50Overhead
- 16:17POBS - + Quick command
- 18:24Empirical Study
- 19:24Design
- 20:42Demo time!
- 23:13Summary
- 24:04Thanks for listening!
related talks
52mChaos Engineering with Spring Boot
39mWhy SREs can't afford to NOT do Chaos Engineering
15mChaos - Does It Need to Be Included in Test Strategy?
19mLearnings from Chaos experiments
34mChaos Engineering For Hybrid Targets With LitmusChaos
47mManaging and monitoring a JVM
37mUse Chaos Testing to Build Resilience in the Software
40mCatching More Bugs with Fewer False Alarms