Our client was a US-based, non-profit organization working within the educational field, striving to ensure equal rights of students in low-income groups across the country.
Our client needed to migrate its database and application system from PostgreSQL to Oracle Exadata. The scope of this migration was extensive, and included data, schema, and fourteen Java applications with Torque Framework and Embedded SQL. All in all, there were more than five thousand scripts and two million lines of code. The challenge in this project, apart from the limited migration time frame, was the necessity of having a specific, customized, and dynamic SQLWays installation for the Torque Framework.
Several competitors of Ispirer from major migration vendors were considered for performing this task. Following a highly competitive tender process, we were selected to provide a full-cycle conversion service using the SQLWays migration tool and post-project 24/7 technical support for three months.
A highly qualified team of database and application migration specialists was assigned to the project, including an on-site consultant.
Agile methodology was applied to this project, which was divided into several sprints. Two teams almost worked in real-time; Ispirer’s technical specialists devoted extensive working hours to the project. Regular stand-up meetings and comprehensive status reporting enabled us to execute the project successfully. These methods, along with on-site consulting, helped us to maximize communication efficiency in spite of minor delays in responses.
Results and Advantages
The project performance required two consolidations because the application and database sources were constantly modified and edited. In the end, all requirements of the customer pertaining to the target code were successfully implemented. One of the major advantages of the proposed solution was the possibility of automating the migration process in order to minimize manual interventions as well as product customization efforts with each database.
Despite an eight-hour downtime to migrate approximately four hundred GB of production data, the team successfully accomplished this task within a short period of time.