Split the MT5 estate into operating roles
Start with a component map rather than a terminal total. Identify live account workspaces, historical-data processing, integration clients and supporting databases. Assign an owner, data location and recovery requirement to each role.
Independent workspaces can be grouped by account purpose and operational responsibility. Keep jobs that produce sustained compute or storage demand identifiable so they can be scheduled, isolated or moved without disrupting live applications.
Budget concurrent work and storage growth
Sizing should reflect which tasks overlap. A historical-data job that is harmless overnight may compete with active terminals during another session. Record both the ordinary load and the combined peak, then agree a capacity reserve for the whole environment.
- Inspect CPU pressure while all intended concurrent jobs are active.
- Measure working memory across terminals and supporting services.
- Record growth of history, logs, caches and retained recovery copies.
- Check sustained storage behaviour when data jobs overlap with terminal operation.
- Plan independent recovery storage rather than filling the primary machine with every copy.
Keep terminal instances identifiable
Install concurrent MT5 copies in separate directories and map their data folders to the appropriate Windows user. Use clear names in the application inventory and record the account-server destination. This is especially useful when several instances use similar chart templates.
| Layer | Document before handover |
|---|---|
| Live workspaces | Installation, data folder, program set and expected account-server connection. |
| Supporting applications | Interfaces, service accounts, dependencies and startup order. |
| Data jobs | Concurrency, schedule, storage paths and resource limits. |
| Recovery environment | Restoration sequence, validation steps and authority to switch workloads. |
The multiple-terminal guide explains the same inventory method for virtual deployments.
Accept the delivered server as a complete system
- Confirm the offered processor, memory, usable storage and network scope.
- Build a representative test environment with the required application groups.
- Run the agreed concurrent workload and record capacity observations.
- Restore a recovery copy and check the service startup sequence.
- Approve the migration window, escalation path and rollback procedure.
If the project requires broker-side platform components, start with MT5 White Label planning and the broker server architecture. A client-terminal environment does not establish platform-side permissions.