The platform running it was a Windows Server foundation well past mainstream support. By 2020, it was carrying 65 FTP and SFTP links across treasury, logistics, product operations, and customer-facing exchanges.

There was no realistic path to keep it. CPU licensing was throttling throughput. There was no encryption. No logging. No monitoring of any kind. Every transfer was a quiet bet that nothing would go wrong this time.

The question wasn't whether to replace it. It was how to replace it without breaking any of the flows the business already ran on.

Challange

The IMTS estate had grown to 65 FTP and SFTP links carrying the kind of traffic a business can't afford to lose - treasury flows, logistics exchanges, product data, customer-facing transfers. Every one of those flows mattered. None of them had the protections you'd expect on infrastructure of that criticality.

There was no encryption on the transfers. No structured logging of who sent what, when. No scalability path to absorb growth without re-licensing. And CPU licensing on the underlying platform was acting as a soft cap on throughput - the system couldn't run faster than its licence allowed, even when the network could handle more.

The risk register read like a list of single points of failure: every flow, every protocol, every business domain. Outages weren't predicted. They were budgeted for.

Solution

The replacement was not a like-for-like swap. IMTS was being asked to do work it couldn't safely do, and a new platform that simply absorbed the same flows would inherit the same throttling.

We took the harder route: rebuild the file-transfer estate on webMethods Active Transfer, but architect around the licensing constraints rather than spending past them. The new platform was designed to extract more throughput from the same hardware footprint - not by upgrading the hardware, but by being smarter about how transfers were structured.

Protocols moved up at the same time. FTP became SFTP and SMB across the business. Where treasury required it, FTPS was retained for banking flows. webMethods Integration Server became the central management plane, giving the operation one place to monitor, troubleshoot, and govern the entire estate.

The cutover was on time and without disruption. Not "with planned downtime" but without disruption.

Result

+50% throughput on the same hardware. The license-aware design extracted more from the existing footprint than the previous architecture could.

65 interfaces became 500+. The platform scaled to absorb roughly eight times its starting capacity without re-architecting.

The estate is cloud-native ready. Object storage connectors for Azure Blob and Amazon S3 are in the platform, available when business decisions call for them.

Full monitoring and secure transport. Every transfer is now observable. Every protocol meets modern security expectations.

What was once a quiet liability - 65 flows running on infrastructure nobody wanted to admit was still in production is now a foundation the business can scale on.