A guest post by J2 account manager, Brad Schuster
When the development team at Surgical Information Systems (SIS) asked if we could help them make some quick-win performance improvements to one of their applications, we jumped at the chance.
Setting the stage
One of the SIS applications for financial and clinical management for ambulatory surgery centers (ASCs) needed a check up. As with all successful data management solutions, SIS faced the common growing pains associated with having to rapidly scale to meet the needs of clients large and small. An uptick in performance-related tickets motivated SIS to reach out to J2 so we could explore ways to reduce data processing time for clients across the board, including enterprise-level clients with truly massive datasets.
Our optimization analysis focused on two potential culprits:
Performance Culprit 1: Hardware setup and configuration
We started by analyzing the hardware setup and configuration for a handful of key sites. By running specialized system-level performance metric gathering tools, we were able to capture and analyze key information related to physical resources, global cache settings, memory heap settings, lock table sizes, and per-process memory limits, among numerous other key performance indicators. From this information and analysis, we were able quickly recommend and implement valuable configuration changes at sites with the most acute performance woes.
Performance Culprit 2: Software under the hood
The next highest priority was a software-level investigation. We spent quality time getting to know the code base and the folks behind it. The team targeted key data processing areas for fast performance wins by shifting the way data is handled both in memory and on disk. We also identified places where UI-side functionality tweaks could potentially have a large impact on user-based data filtering, in the form of improved user experience and faster report processing times.
Results
SIS is currently implementing J2’s recommendations, and the early returns are very positive.
In short order, SIS reported a 50% decrease in performance-related ticket volume just by addressing some of the low hanging fruit, with further reductions expected down the road as they make some of the more effort-intensive modifications. But even before those additional gains are realized, this has truly proven to be a case where a little optimization can make a huge difference!
