Aglet

Triage an API dependency outage fallback

A failed dependency call can affect reads, writes, and reconciliation differently. Freeze one outage-shaped request and record failure class, dependency endpoint, local intent, fallback label, and recovery state. Keep timeout, rejection, rate response, configuration error, and partial outage separate. Keep the dependency label beside local state.

Establish what is happening

  1. Capture the failed dependency

    Record operation, dependency label, environment, failure class, request time, local state, and user-visible result without sensitive payloads. Note whether the request could have produced a remote side effect. Preserve correlation for later recovery.

  2. Compare fallback behavior

    Run a healthy response, temporary failure, explicit rejection, stale fallback, and no-fallback case in a controlled fixture. Compare freshness label, local decision, and reconciliation marker. A value returned from cache must not look current by default.

  3. Bound outage reach

    Group failures by dependency, operation, caller, environment, fallback state, and duration. Separate shared outage, local configuration, rate response, and single-resource failure. If recovery cannot be observed, record the boundary rather than assume the cache is safe.

What to carry forward

Triage ends with one failed dependency call, local intent, fallback or pending state, affected cohort, and recovery gap. Stop unsafe writes and defaults. Route transport, fallback, and reconciliation questions separately. Keep unknown work pending until recovery evidence exists.

Technical background: RFC Editor reference.

Keep the decision with the work.

Use a Work Item in Aglet to record the problem, the evidence you have, and the next decision. Add an owner and priority, then keep updates in the discussion so the next person can follow the reasoning.

Create an account See the product workflow