Software Engineer at John Deere
Jan-2013 to Mar-2014
Functionality – This project re-engineered the process to store JDLink data (Modular Telematics Gateway sending machine health information). This data comprised of machine measurements, alerts, and events collected on a 24/7 basis. This process is a high-volume, “big data” batch cycle that can pass up to 200 million records a night. A new database node strategy was devised to allow machine data to be spread across multiple, smaller database instances. Single code multiple runs data model was enabled to migrate current Z/OS JDLink Data Warehouse and Mart to Netezza. Fully automated process with near to real time data leading to time reduction from 12-24 hours to four times daily execution. Redesign facilitated more than 50% performance improvement in the application. Highlights – Push Down Optimization, Netezza Bulk Writer, Netezza Geospatial Store, MSRMT_BASE - 60 billion records, Resource throttle, Dynamic creation of parameter file, one click report, Clone tables, Concurrent work