Practice vocabulary for learning from chaos experiments: unknown dependencies, unexpected behavior, discovering gaps, and actionable insights.
0 / 5 completed
1 / 5
A post-experiment summary says 'The experiment revealed an unknown dependency.' What is an unknown dependency in this context?
Unknown dependencies are often the most valuable discoveries from chaos experiments — services, databases, or external calls that the system silently depends on, but that no team member was aware of. Discovering them before an outage is the point.
2 / 5
Your experiment report says 'The system behaved unexpectedly when the cache was injected with latency.' What should the team do with this finding?
Unexpected behavior revealed by chaos experiments is a finding — it means the system has a gap in its resilience assumptions. The team should document it, understand the root cause, and address it (e.g., add timeouts, fallbacks, or circuit breakers).
3 / 5
Your team says 'We discovered the backup wasn't working until the chaos test.' What type of gap did the chaos experiment surface?
This is a critical class of chaos finding: an assumed recovery mechanism that was never actually tested. Chaos engineering frequently reveals that backups, failover procedures, or circuit breakers exist in theory but don't work in practice.
4 / 5
The experiment learning statement reads: 'We need circuit breakers on this call.' What prompted this conclusion?
Circuit breakers are a resilience pattern that stops requests to a failing service after a threshold, preventing cascade failures. Chaos experiments that induce downstream failures reveal exactly where circuit breakers are missing or not configured correctly.
5 / 5
A chaos retrospective uses the phrase 'the learning was: the system degrades silently without alerting anyone.' What action should this trigger?
Silent degradation is a dangerous failure mode — the system is failing but no one knows. The chaos finding should directly drive adding observability (metrics, alerts, dashboards) so the failure mode is immediately visible if it occurs in production.