BMC Case Studies Seeing is Achieving: How Visualization Helps a Major European Bank Manage More Mainframe Code Better, Faster and with Fewer People
Edit This Case Study Record
BMC Logo

Seeing is Achieving: How Visualization Helps a Major European Bank Manage More Mainframe Code Better, Faster and with Fewer People

BMC
Analytics & Modeling - Process Analytics
Application Infrastructure & Middleware - Data Exchange & Integration
Finance & Insurance
Discrete Manufacturing
Quality Assurance
Predictive Maintenance
Process Control & Optimization
Software Design & Engineering Services
System Integration
The bank, one of Europe's largest private banks, is facing a common IT management challenge. The bank has stopped developing new mainframe applications and allowed its mainframe development staff to shrink by about 30% through attrition. However, the bank's mainframe development needs are increasing. The use of the bank's mainframe applications and data by new and existing distributed, web, and mobile applications is increasing. Developers no longer build new COBOL applications but must constantly modify mainframe code to assist non-mainframe applications. The bank can't wait forever for coding changes to be completed. Competition in the financial services market is intense, making time-to-benefit important, whether for customers' mobile experiences or new capabilities for staff, and putting pressures on the mainframe software development lifecycle. That lifecycle has to produce defect-free code. The bank's mainframe applications are critical to everything it does and everything customers and regulators expect. Mainframe test/QA must be rigorous and fast. If the testing process uncovers a problem, it must be rapidly and accurately diagnosed, fixed, and tested again.
Read More
The customer is one of Europe's largest private banks. The bank has stopped developing new mainframe applications and allowed its mainframe development staff to shrink by about 30% through attrition. However, the bank's mainframe development needs are increasing. The use of the bank's mainframe applications and data by new and existing distributed, web, and mobile applications is increasing. Developers no longer build new COBOL applications but must constantly modify mainframe code to assist non-mainframe applications. The bank's mainframe applications are critical to everything it does and everything customers and regulators expect. Mainframe test/QA must be rigorous and fast. If the testing process uncovers a problem, it must be rapidly and accurately diagnosed, fixed, and tested again.
Read More
The solution, according to R.C., has been embracing visualization with Compuware tools like Topaz, Xpediter, File-AID, and Abend-AID, providing developers with visually intuitive insight into application relationships, runtime behaviors, and problems in programming logic flows. Issues that previously took hours to find now take developers minutes with these visualization tools, which also help developers understand in advance the implications of coding changes to avoid re-coding and re-testing. In one example, given the ability to graphically visualize a mainframe application’s runtime activity in high-level production and “drill down” to a detailed view, R.C.’s team noticed a batch process was completing later and later. If the trend continued, the process would not complete before the open of business, spelling trouble for the bank. Using Compuware’s Topaz for Program Analysis, R.C.’s team zeroed in on a sort-program call repeating hundreds of thousands of times a night. A simple code fix reduced that egregious performance bottleneck to a single call.
Read More
Reduced time to discover and resolve issues in application relationships, runtime behaviors, and program logic flows.
Clarified impact of code changes in advance to avoid re-coding and re-testing.
Reduced batch processing time after visualization of application’s runtime activity in high-level production.
Reduced mainframe development staff by about 30% through attrition.
Issues that previously took hours to find now take developers minutes with visualization tools.
A simple code fix reduced a performance bottleneck from hundreds of thousands of calls to a single call.
Download PDF Version
test test