Hypothesis-driven chaos experiment on a small multi-service application

State a steady-state metric and hypothesis, inject instance loss or added latency with a limited blast radius and abort condition, and report what the fallbacks actually did.

The exercise happens on Self-directed project (chaos engineering principles). Canonical link: https://principlesofchaos.org/. Link check: link responded (5 Oct 2026). Opened from this app 0 times; opening is not counted as an attempt.

Supports: Resilience-evaluation practice relevant to cloud research and reliability-minded systems roles.