Debug cancel-trial versus cancel-plan false hits

Map failure classes for POST /v1/answer: retryable vs fatal vs needs-human. Retries need budgets, jitter, and idempotency rules aligned to cache key includes tenant_id + policy_version; similarity ≥ 0.92 and risk≠high. The chapter’s signature failure — cross-tenant near-duplicate question returns Acme's private pricing blurb — must take a deliberate branch, not a generic catch-all.

Implement the degrade/rollback/refuse behavior cost owner cutting duplicate GPT calls for Acme Corp FAQs needs when CACHE-XTENANT-2 repeats. Prefer scoped controls (one flag, one weight, one tenant, one secret version) over fleet-wide restarts. Preserve evidence; do not delete logs to “clean the demo.”

After containment, check hit_rate ∈ [0.25,0.55] and cross_tenant_hits == 0 moves in the safe direction and watch for retry amplification. Write the stop condition that ends the incident response for this lab.

Chaos note: inject only one fault class at a time and restore fixtures after. Watch for dual failures — dependency down and retry amplifier — which is how cross-tenant near-duplicate question returns Acme's private pricing blurb becomes an outage. Customer communication templates (even if only for the drill) beat silence. If you queue deferred work, define poison-message handling. Budget documents should state the maximum extra spend allowed during retries. Close the loop by linking the containment action to a dashboard panel for hit_rate ∈ [0.25,0.55] and cross_tenant_hits == 0.

Draw a state diagram for degrade modes and put it in the repo as ASCII if needed. Cap concurrent retries across the process, not only per request. Ensure cancellation propagates to downstream HTTP clients. When failing closed, choose a user-visible message that does not leak internals. Practice the single command that flips the kill switch or weight to zero. After recovery, drain or inspect deferred work before declaring green. In this chapter the product is tenant-scoped semantic cache for repeated FAQ paraphrases, the human stakeholder is cost owner cutting duplicate GPT calls for Acme Corp FAQs, and the incident id you design against is CACHE-XTENANT-2. Re-state the oracle in your notes — paraphrase hit returns cached answer with cache=HIT; different tenant never hits — and keep the invariant visible: cache key includes tenant_id + policy_version; similarity ≥ 0.92 and risk≠high. Track hit_rate ∈ [0.25,0.55] and cross_tenant_hits == 0 as the scoreboard. Surface under change control: POST /v1/answer. If you only have forty minutes, finish the fixture for cross-tenant near-duplicate question returns Acme's private pricing blurb before polishing UI. Promotion language stays ternary: promote, hold, or roll back based on evidence, not hope.