Aglet

How to verify one logical operation is processed once

Verification must show that repeated attempts for one logical operation produce one intended effect while distinct operations still proceed independently. Exercise concurrency, timeout, retry, and replay conditions, then inspect durable state, responses, audit history, and recovery after each attempt under representative timing.

Check whether the outcome improved

  1. Replay the original duplicate trigger

    Use the same operation identity, route or event shape, account class, timing, and failure boundary from the baseline. Run concurrent or repeated attempts and record claim, effect, commit, response, and final state. Compare stable identifiers and counts rather than relying on a success status.

  2. Check retries, replay, and recovery

    Retry after a timeout, replay an accepted event, and repeat after a failure before or after the effect boundary. Confirm one logical outcome, clear response ownership, and safe recovery. Verify partial or failed attempts do not leave a second effect waiting for manual cleanup.

  3. Prove distinct operations remain distinct

    Run two operations with different identities but similar input, timing, and target data. Confirm each creates its allowed effect and neither is swallowed by the deduplication rule. Check ordering, audit records, validation, permissions, and customer-visible state for unintended coupling.

What to carry forward

Classify duplicate handling as safe, regressed, unchanged, or inconclusive with interleaving, effect, response, and distinct-operation evidence attached. It passes only when one logical operation has one intended outcome and valid separate operations retain their independent behavior.

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