How should NRUNO handle maintenance without losing or duplicating signals?

What this means in practice

Maintenance needs a defined intake/execution policy: drain or pause, durably persist accepted commands, stop at a safe boundary, reconcile after restart and keep retries idempotent. This page is specifically about “How should NRUNO handle maintenance without losing or duplicating signals?”, so each scenario below is explained by its own mechanism instead of sharing one generic diagnosis.

Real-world scenarios

Scenario A — Deployment

Define maintenance around already accepted work: drain or stop intake deliberately, persist command identity, restart, reconcile cTrader state and resume idempotently so accepted commands neither vanish nor execute twice. For Scenario A — Deployment on question 97, use that evidence specifically to answer “How should NRUNO handle maintenance without losing or duplicating signals?”; keep it separate from the evidence for the other scenarios on this page.

Scenario B — Unexpected restart

Define maintenance around already accepted work: drain or stop intake deliberately, persist command identity, restart, reconcile cTrader state and resume idempotently so accepted commands neither vanish nor execute twice. For Scenario B — Unexpected restart on question 97, use that evidence specifically to answer “How should NRUNO handle maintenance without losing or duplicating signals?”; keep it separate from the evidence for the other scenarios on this page.

Scenario C — Accepted-before-maintenance

Define maintenance around already accepted work: drain or stop intake deliberately, persist command identity, restart, reconcile cTrader state and resume idempotently so accepted commands neither vanish nor execute twice. For Scenario C — Accepted-before-maintenance on question 97, use that evidence specifically to answer “How should NRUNO handle maintenance without losing or duplicating signals?”; keep it separate from the evidence for the other scenarios on this page.

What to check

• intended signal/action
• last stage that definitely succeeded
• first stage that differs from intent
• final cTrader state after the event

Practical rule

For “How should NRUNO handle maintenance without losing or duplicating signals?”, change only the first layer whose evidence no longer matches the intended action. Preserve signal identity, timestamps and final cTrader state, and reproduce execution-affecting changes on demo before live use.