Managed File Transfer

Suppose a retailer collects completed sales exports from a partner's SFTP server and delivers them to an internal reporting application each morning. IT operates several such feeds and needs predictable timing, a record of what transferred, and a procedure for late or rejected data.

Define the Delivery Contract

Agree with the producer on the pickup path, filename convention, file format, expected volume, and time the export becomes ready. Use unique delivery filenames when possible. The producer should finish writing before exposing a file for pickup and retain it long enough to cover a failed collection.

Agree with the reporting team on the destination path and the acknowledgment that means an export was successfully imported. A transfer completion proves a different stage from a successful report refresh.

Record how corrections and duplicates are identified. If a corrected export reuses the original filename, the transfer mechanism and the receiving application both need defined replacement behavior.

Group the Resources

Place the feed's Remote Servers, folders, Syncs, and monitoring configuration in a Reporting Workspace if that grouping fits the organization. The Workspace may contain several related feeds or just this one. The same IT team can operate it alongside payroll and supplier Workspaces.

For testing, use a separate Workspace when the experiment concerns those resources alone. Use a Child Site when independent site settings must be tested. In both cases, use test endpoints and credentials; the environment's name does not prevent a connection from reaching production.

Create the SFTP Remote Server with an account scoped to the pickup directory. Connect the internal destination through the appropriate Remote Server type, including the Agent when it supplies access to an internal server.

Build the Transfer in Stages

In this example, an inbound Sync collects files into Files.com native storage. Keeping a received copy makes investigation and replay possible while the remote source is unavailable. A separate outbound transfer delivers the files to the reporting application's pickup folder.

Use Sync when one-way replication fits each stage. Sync compares names and sizes, so same-name, same-size corrections require another approach. Use an Automation with appropriate selection and replacement settings when files need renaming, routing, or different overwrite behavior.

Two independently scheduled stages are not a dependency chain. Allow for late collection and define when a file becomes eligible for onward delivery. If content must be validated before release, keep received files outside the delivery source until your validation process explicitly releases them. An Expectation can report a validation failure; creating one does not automatically block another transfer.

If the process needs external transformation software, let that software own its processing and publish only completed output to the delivery stage. File Processing & Routing explains the alternatives. Automations v2 provides connected multi-step workflows, but remains an opt-in early beta and is not recommended for business-critical production workflows.

Operate Against the Deadline

Set schedules in the intended time zone and leave time for transfer, import, and recovery before the report deadline. Check start-window and overlap behavior rather than treating a configured time as an exact completion promise.

Use Expectations to detect missing or invalid inputs and Event Channels to deliver operational outcomes to IT. Inspect partial failures per file. For a destination using buffered uploads, verify onward delivery rather than relying only on the incoming upload completion.

Keep the receiving application's import acknowledgment alongside the transfer evidence. Assign the producer's team to missing exports, IT to connection and delivery failures, and the reporting team to import rejections.

Prove Recovery and Cut Over

Test a representative export, an unavailable endpoint, a malformed file, and repeated delivery. Confirm whether retrying creates another business transaction or replaces an existing one. Establish how an operator stops delivery while preserving the received input.

During an MFT migration, choose a cutover point with the old system's owner so both systems do not process the same export unintentionally. Verify a controlled first production delivery and its import result. Keep the previous process recoverable until the new flow has met the agreed acceptance criteria, following Production Validation & Change Control.