Stories from the Front Lines of Complex Workload Management: Reducing Costs & Increasing Visibility with Terma’s Solution
March 14, 2019 1:17 pm
For more than ten years, we here at Terma Software have been down in the trenches of Workload Automation. Workload environments have become increasingly more complex, and we have been developing our intelligent workload analytics solution right alongside them. Our subject matter expertise spans across multiple vendors and platforms, and various configurations of the two. If there’s something that could go wrong with the management of a job scheduling system, then we’ve seen it before. Here are just a few examples of the ways we’ve been able to help our customers significantly improve their workload management.
Workload outages can result in business disruptions that directly impact the bottom line. A failure could, for example, prevent a financial firm from filing the necessary reports to open their trading desks for business in the morning. We’ve witnessed multiple situations similar to this where our solution enabled our customers to avoid costly issues:
- When a large global bank experienced an unexpected computer network shutdown, they were fined approximately $20 million by the government regulatory agency. We were able to provide the proper controls and processes the bank needed to prevent something like that from happening again.
- After implementing our solution, a large global financial company realized a quick ROI by identifying areas where workload performance was naturally degrading over time. The company was able to address these instances before they affected SLAs.
- Another large financial firm that implemented our AIOps platform saw an immediate ROI by reducing batch inefficiencies and shortening the time needed to fix batch issues, which avoided SLA violations and fines.
One of the major issues with the increasing complexity of Workload Automation is a lack of visibility into workload environments. When companies are using multiple legacy scheduling systems and/or adding on in-house developed platforms, there is no way to view jobstreams from end-to-end or to accurately discern the interdependencies. The following two companies were able to vastly increase their understanding of their workload processing, which led to better decision-making and more effective management:
- A large bank was using a mainframe system to run 1 million jobs per night, but it also had a tightly connected distributed system. Jobs running on the mainframe triggered jobs on the distributed system and vice versa. However, the bank had no way to see across all processes to know exactly how those jobs were connected. By implementing our solution, the bank gained that visibility and was then able to make informed decisions about the order and processing of workloads to avert potential problems.
- A multinational investment bank with a complex workload environment was running large volumes of critical jobs nightly. While most of these jobs ran on a mainframe, about 40,000 per night ran on a homegrown distributed solution. The two systems had processes that were dependent on each other, but the bank was unable to view the intricacies. We were able to provide a REST API in addition to our workload analytics that created visibility into the job scheduling system and the ability to model relationships between jobs and between platforms. The bank could then clearly see the critical paths across all platforms and applications. This holistic view combined with our predictive capabilities enabled the bank to prevent workload issues before they occured.