Shadow Testing
Updated 2026-08-10
INTRODUCTION
English translation pending.
CORE DEFINITION
Shadow testing duplicates live production traffic and routes it into the new version of a system in parallel with the existing one. The shadow system executes its logic but its responses are discarded, so users are unaffected while the new version is exercised against real data and real load. Differences between the two systems are compared before any traffic is switched, which makes it a low-risk way to validate a replacement under conditions a test environment cannot reproduce.
SCAFFOLDING EFFECT
Reduce cognitive load
- Rehearse in parallel: run the new approach alongside the current one without letting it take effect. - Compare outputs: check where the two diverge on real inputs before switching anything over. - Build experience cheaply: simulate decisions in advance and compare them with what actually happened.
Anchor fast decisions
A test environment approximates production, so it misses the data distributions, load patterns and edge cases that only real traffic contains. Replaying live traffic through the new system exercises exactly those conditions while the shadow's output is ignored, so any failure has no user impact. Divergence between the two systems becomes the evidence for whether the replacement is ready.
MINIMUM ACTION
In progress 0/1Practice this model in one real situation:
account_treeGenealogyexpand_more
menu_bookReferencesexpand_more
Source support: Explicit
- istio.iohttps://istio.io/latest/docs/tasks/traffic-management/mirroring/verified
PRIVATE NOTES · Only visible to you
SAVED Q&A
ENTRY Q&A · Private saving available
Ask with a clear boundary
thinkingmodels answers from published entry context only.
Your question is sent to thinkingmodels. The answer uses public entry context only.
RELATED MODELS