Choose an outage scenario and its boundary
| Item | Question to answer |
|---|---|
| Failure | Which component, location or dependency is unavailable? |
| Data state | Which usable record or configuration state can be recovered? |
| Service priority | Which workflows need to return first? |
| Decision authority | Who declares recovery and approves a return to normal operation? |
Define the recovery-time and recovery-point objectives for that scenario. Keep objectives distinct from tested results: a desired target is not evidence that the current environment meets it.
Prepare the dependencies needed for a clean restore
Inventory software versions, approved installation sources, platform configuration, protected credentials, network rules, certificates and external connection requirements. Confirm the permitted restoration method with the relevant software and service providers. A file copy is not a universal method of reconstructing a platform environment.
Use an isolated or otherwise approved exercise environment. Establish how it is prevented from unintentionally sending production transactions or conflicting with live services. Agree the exercise boundary before any change to live routing.
Run, reconcile and improve the procedure
Follow the written steps, recording elapsed stages, manual dependencies and unexpected decisions. Validate access, the agreed data state, representative account workflows and the required external connections. Reconcile records using the process authorised for the environment; reaching a login screen is not sufficient acceptance.
After the exercise, record what worked, what remained unavailable and which assumptions failed. Assign corrective actions and repeat the affected part after changes. Use the MT5 recovery project scope when the team needs help turning these findings into a defined implementation brief.