A large pharmaceutical company invited Full360 to migrate their Oracle data warehouse from their data center to the Amazon Cloud. We selected Vertica as the database of choice and performed a successful migration end to end.

The Problem

The client desired to upgrade the capacity and performance of their Oracle DW. The licensing costs were prohibitive, and their staff and third-party staff augmentation team had no cloud experience. The applications supported by the Oracle DW was built over the course of 5 years by 20 DBAs and developers. With over 200 data feeds needing transforms including sftp from vendors, Salesforce extracts, Google analytics and market analysis data, this was considered a very complex set of applications to migrate. The arcane complexities of configuring Oracle at the limits of its capacity were costly to debug and causing divergence from best practices and simplicity.

The Solution

Full 360 created custom producers for every unique data-feed and built a strucutured data lake. An integrated UI was built to control and schedule bulk uploads for all file clases, including built in flexibility to accomodate new feeds. All metadata for workflow control was built into the UI back-ended by DynamoDB. A three node Vertica cluster was installed into the cloud environment custom built by Full 360.

The Results

Working with staff and consultants based on the East Cost, the Full 360 team migrated the entire data warehouse that included financial, marketing, sales and inventory data in under six months.

• Enterprise IT Simplified Within weeks of the implementation, the drama and tension previously associated with managing an Enterprise-wide Oracle Data Warehouse vanished. As time went on, people recognized that it didn’t have to be such a pain. This kick-started DevOps practices at the customer.
• Size Managed There were over 200 sources of data that landed in 24 different schemas using more than 2500 tables. This was a set of legacy applications with many thousands of lines of SQL maintained by a large staff of DBAs and Oracle developers.
• Complexity Handled Using Amazon DBM and Full 360 SneaQL, our team was able to convert all code including hundreds of PL/SQL stored procedures. Not only did we do a lift and shift, but added much-needed performance, scalability and lowered costs as well.
• Performance Improved The resulting data warehouse, powered by a three node cluster of Vertica database nodes increased overall load and query performance. By containerizing SneaQL, maintenance and business rule changes were made easier and more flexible resulting in a lower time to development. Disaster recovery was simplified and with a cloud-based data volume license operating costs were reduced. No more per processor costs.
• Extended Functionality Client was able to quickly add new metrics and analytic capabilities to their data warehouse.
• Support & Managed Services Provided Full 360 integrated with their ticketing system and gradually migrated all technical support from the internal staff and third parties to Full 360 handling the requirements of the SLA, including Disaster Recovery and ‘round the clock support.