Global Fintech Leader Integrates Mainframe into Splunk® with Ironstream

Fintech leader relies on mainframes for unparalleled transaction processing power

A large financial services company had become a leading provider of core banking services that underpin credit card transaction processing for millions of consumers and businesses worldwide. Like many major corporations, the company had relied for decades on massively powerful mainframe systems to run its business.

But today, no corporation of size can compete without extensive use of modern IT operations management platforms that provide real-time, enterprise-wide visibility into what’s going on across their infrastructure. The company invested in Splunk to get this single comprehensive view. However, when it wanted to extend Splunk’s world-class analytics capabilities to its mainframe systems, it quickly became apparent this would be no easy task. The complexities of mainframe architecture, data formats, and programming made it all but impossible to achieve with in-house efforts. It needed to find a more workable solution.

The firm’s core transaction systems rely on the IBM Z platform because of its unequaled power and speed for batch and transaction processing. Real-time approvals of credit card transactions require ultra-fast processing, from ensuring available balances to confirming CVV security codes, comparing normal account behaviors, checking against fraud detection systems, and much more. And these checks must be completed within milliseconds per authorization transaction to ensure that customers and merchants are not left waiting.

In addition to such real-time services, many batch processes must be run nightly against millions of accounts. These processes generate monthly statements, calculate interest and minimum payments, create new credit cards and pins, account behavior scoring, and much more.

Mainframe performance is robust and comprehensive. But it’s native monitoring, alerting, and reporting can be less impressive. It tends to be on-platform, siloed, the realm of subject matter experts, and out-of-step with modern ITOM solutions such as Splunk.

Ironstream automates the collection and forwarding of critical mainframe logs

The company attempted to develop custom mainframe programs for monitoring its applications, but those efforts required many months of work. Even then, because they had to run in tandem with production transaction processing, gathering
of information added significant costly CPU overhead.

Making things worse, the interface screens for these solutions only displayed results in real-time and had to be physically
monitored, with eyes on the console. So, these custom applications were only run when absolutely necessary to investigate critical issues and then shut down.

Before long, the company concluded that the mainframe is extremely good at providing transaction services but not nearly
as good at providing information about what is taking place right now or what may happen in the near future. In short, its
mainframe systems couldn’t provide the level of real-time visibility and business insights that were needed by a wider audience.