Accept

By using this web site you accept our use of cookies.

More information about cookies

EVRY India Optimizes a Retail Chain’s Performance

EVRY India was engaged in providing a solution for a retail customer, an American multinational corporation that runs a chain of large discount department stores. The customer faced issues such as dissatisfactory results obtained through BI reports, high resource consumption such as CPU & input/output, and high transaction response time.

 

Challenges

The challenges in delivering an ideal solution to our customer involved – designing a single global Enterprise Data Warehouse design from multiple source systems’ data, enablement of multi-currency support, multi-lingual systems & spread over different geographic locations, accommodating multiple financial years of different countries in a single model, minimal support from the business team, inaccurate or non-availability of documentation, and unavailability of the engineering team during the design phase.

 

Solution

  • EVRY India took a top-down tuning methodology that was applied to a performance problem. The diagnostics were collected using a Teradata viewpoint. They were analyzed for various critical parameters like transaction response time and resource usage such as CPU and I/O. In addition, we also analyzed the database and report design layers to identify the performance bottlenecks. We identified the performance hindering queries from the collected data.
  • We presented various ways of database and reports optimization as well as verification that the recommended solution solved the problem. It summarizes a few key solutions that were implemented for performance improvement.
  • Added Partition Primary Index (PPI) and Secondary Indexes (SI) reduced resource consumption such as I/O and CPU.
  • The JIs (Join Indexes) and AJIs (Aggregated Join Index) appropriately resulted in better execution of our plan, response time and improved savings in CPU and I/O.

All the above solutions were implemented against the test database, which resulted in enhanced query response time and resource consumption. The solution implementation was complete before deploying them into the production environment. Further, the production data with test database was synchronized using Teracopy and loading utilities. Also, we identified and redesigned the PI (Primary Indexes) to reduce HOT AMP operation.

 

Result

The customer benefited tremendously with our solution through better transaction response time (70% improvement in response time), up to 70% improvement in CPU & I/O utilization, quick turn-around time for its Business Intelligence reports, increase in end-user productivity & efficiency in decision making, greater acceptability in using the reports by end users in decision making, increased tasks loading capacity, augmented number of concurrent users & queries,  reduced wait time for users, and decreased resource utilization, CPU consumption & I/O costs.


 

About the Customer

The retail customer, along with operating an extensive chain of department stores also maintains a chain of warehouse stores. The customer runs more than 8000 stores across the world. Its hassle-free merchandising division ensures that products are constantly available to consumers across all its retail outlets, primarily through planning, stocking shelves and displays.