When one thinks of data access issues, “big business” and “the enterprise” often spring to mind first. While these issues are common business problems, this does not mean that other industries are immune to them. This was illustrated clearly in DataDirect’s recent work with the Arkansas Department of Education (ADE). The department needed to provide “on demand” access to data such as test scores, attendance, graduation rates and the like from school districts across the state – data that up until then was only transferred nine times a year. The data expansion project was part of ADE’s goal of improving educational results across the board to better prepare Arkansas students for success in college and beyond.
Though the system had been capable of handling data pulls from all of the districts nine times per year, the addition of nightly and on-demand data pulls required significantly more processing capabilities and swamped the system. Though the updated system worked well with pilot school districts, the built-in database drivers failed to perform once data was pulled from all districts. Furthermore, the built-in drivers were failing with read errors when executed against the full data set.
DataDirect’s ODBC drivers, however, were able to resolve the load errors and delays. The drivers enable successful extraction, transformation, and load (ETL) processing for the nightly database pulls from the districts. And, the whole process works 20% faster than it had in comparison to the earlier, less frequent cycle of loading. The gain in performance helped the ADE avoid an investment in additional processing capacity and reduce errors during ETL, which translated into more efficient management of the entire system and money saved for reinvestment in teachers, technology and other resources. For more details on the challenges faced, the deployment and the results, check out the full case study. The study went live a couple of weeks ago, just in time for the 20th anniversary of ODBC (check out our other post about that). Just another example of how after twenty years, the ODBC standard is still adding value, kicking ass and taking names across sectors!