Workload-Centric Solution Saves Supermarket Chain Money – Today & Tomorrow
With Elevate DB2, the company identified 50% of Db2 CPU savings
Db2 processing at this grocery wholesaler and supermarket chain was taking up more than half the mainframe’s CPU capacity. At the same time, the company’s transaction volume was growing. A costly mainframe upgrade was in the cards unless the company could find a way to bring the Db2 workload under better control— in other words, unless a certain amount of CPU capacity could be recovered from the ongoing Db2 work. The Db2 tools then in use, however, had proved to be not up to the task of dialing back the appetite for CPU time. Switching Db2 tools was the only option, but the company needed to find the right way to do it.
The supermarket chain initially engaged with a small outside group to supplement its busy DBA staff and for some stop-gap tuning efforts. Then, after a marketplace survey of Db2 tuning tools, the company chose Syncsort Optimize DB2 as the solution, having concluded that a workload-centric approach to optimization — the approach that is unique to Syncsort Optimize DB2 — was the best choice for identifying Db2 workload savings opportunities.
The company turned to Syncsort Optimize DB2 as the solution, having concluded that a workload-centric approach to optimization — the approach that is unique to Syncsort Optimize DB2 — was the best choice for identifying Db2 workload savings opportunities.
Upon acquisition, Syncsort Optimize DB2:
- Installed in less than three hours, and the first useful results were produced the same day.
- Provided the ability to consolidate and then to rank functionally identical SQLs, ensuring immediate understanding of the impact on resource consumption of each group of functionally identical SQLs.
- Made CPU capacity recovery easier by giving the DBAs the ability to run what- if cost projections of changes in SQL syntax and indexing.
- Identified where each target statement was used across the application portfolio, ensuring that all instances of a target SQL were modified at the same time.