Instrument metrics and rollout decisions

Instrument support-answer service comparing retriever_v1 vs reranking retriever_v2 so experiment owner measuring groundedness lift with SRM checks can answer: demand, errors, latency/age, saturation. Emit fields needed by groundedness_lift_pp and srm_pvalue with bounded labels. Sample successful high-volume traces; keep errors and rollout transitions denser within policy.

Define at least one alert that would fire for AB-SRM-FAIL-27, with a for/pending window that survives deploy blips. Missing scrape or missing revision labels must not look like health. Include a trace/log example id format you will actually search.

Add a unit/integration check that metrics increment on the happy path and on the reassignment every request → users flicker; metrics uninterpretable branch. Store machine-readable output in CI artifacts when practical.

Cardinality discipline: tenant and revision are usually enough; raw question text is not a label. Exemplars or trace links beat screenshots alone when debugging AB-SRM-FAIL-27. Define who owns alert fatigue review. If you export to a vendor, record retention and access. Synthetic probes should use non-sensitive fixtures and still exercise POST /v1/support/answer. Practice the query you will type at 2am once, while calm.

Document the exact PromQL or log query in the runbook stub for this service. Verify histograms have buckets around your SLO target. Add a canary synthetic check that exercises the oracle path every few minutes in staging. Confirm that PII redaction happens before export. Track build/version as a label on the golden signals. Delete noisy debug logs before they become accidental product dependencies. In this chapter the product is support-answer service comparing retriever_v1 vs reranking retriever_v2, the human stakeholder is experiment owner measuring groundedness lift with SRM checks, and the incident id you design against is AB-SRM-FAIL-27. Re-state the oracle in your notes — after 20k sessions, v2 groundedness +2.1pp, p95 latency +80ms within budget — and keep the invariant visible: assignment sticky by user_id; analysis gated on sample ratio mismatch (SRM) p>0.001 fail. Track groundedness_lift_pp and srm_pvalue as the scoreboard. Surface under change control: POST /v1/support/answer. If you only have forty minutes, finish the fixture for reassignment every request → users flicker; metrics uninterpretable before polishing UI. Promotion language stays ternary: promote, hold, or roll back based on evidence, not hope.

Set a timer for twelve minutes and attempt only the observability concerns for A/B test lab. Speak aloud what experiment owner measuring groundedness lift with SRM checks would see on success versus on reassignment every request → users flicker; metrics uninterpretable. Write three bullet artifacts you must keep: a fixture name tied to AB-SRM-FAIL-27, a metric query for groundedness_lift_pp and srm_pvalue, and the rollback or refuse command for POST /v1/support/answer. Then extend the notes with one sentence on how assignment sticky by user_id; analysis gated on sample ratio mismatch (SRM) p>0.001 fail would be violated if you skipped this page. Close by restating the oracle in your own words without looking: after 20k sessions, v2 groundedness +2.1pp, p95 latency +80ms within budget.