Introduction
At a leading company specializing in the production of eyeglass and sunglass frames, Blue BI managed a system migration project from IBM Netezza to SAP HANA.
Initial Context
The company utilized several DWHs built on IBM Netezza to support reporting systems for analyzing and managing data across multiple sales channels (wholesale, retail, and e-commerce). These data were exposed and analyzed using front-end tools such as SAP BO (Business Objects) and QlikView. SAP Dataservices was used for ETL data management, with SAP being the primary ERP system.
The company lacked a unified DWH encompassing all sales data, relying instead on a set of reports, specifically a QlikView dashboard, which integrated data from various channels (wholesale, retail, and e-commerce) to provide a global yet highly aggregated view. The main challenge of this reporting was data integration to support cross-channel analysis.
Objectives
The project had multiple objectives:
- Transition from using Netezza DWH and Dataservices to SAP HANA
- Reduce reliance on physical resources in favor of virtual one
- Normalize data at the back-end level rather than the front-end level
- Eliminate outdated integration logic and streamline current processes.
- Improve key reporting systems and accessory dashboards
- Centralize data and generate cross-channel reporting.
Migration to SAP HANA
Strategy
The implementation of SAP HANA as a new technology aligned with the objective of reducing physical resources in favor of virtual solutions. The strategy involved replacing the physical steps typical of SAP Dataservices (workflows, dataflows, temporary table construction) with virtual steps using SAP HANA Calculation Views.
To ensure continuity in data access for users, a minimal impact approach was adopted for the front-end layer. Calculation Views were implemented to maintain structures similar to those existing on IBM Netezza, leveraging Blue BI’s expertise to deliver a winning strategy.
Preliminary Phase
The project began with an assessment of the existing logic, resulting in the revision and simplification of certain Dataservices steps. This aimed to stabilize and optimize the current environment before proceeding with the migration.
This phase was successful, as performance improvements were achieved in the existing architecture, thanks to Blue BI’s cross-technology expertise. Similar optimizations were applied to QlikView, refining data models and dashboard scripts to streamline and update data loading processes, including the generation of proprietary QVD files.
First Migration Step
The next phase focused on implementing SAP HANA through successive steps, starting with specific master data. Several Calculation Views were created on SAP HANA to replace Dataservices flows feeding these master data.
In some cases, the activities involved not only “migration” but also “integration,” consolidating multiple Dataservices flows into a single Calculation View (CV). For example, a single article master data view was created on SAP HANA, integrating data from wholesale, retail, and e-commerce channels to provide comprehensive and consistent data for analysis tools, avoiding the need for such integrations within the tools themselves.
Second Migration Step
After building master data on SAP HANA, a parallel run period was conducted with Netezza-Dataservices, enabling a gradual switch of reports and dashboards to the new architecture.
The next project step focused on migrating fact tables. For performance and business reasons, separate CVs were implemented instead of consolidating typically separate domains (wholesale, retail, e-commerce, etc.) into a single structure.
Performance Tuning
Following the development of CVs on SAP HANA, a tuning activity improved performance measured during implementation. This involved steps like materialization within the HANA environment, balancing virtual and physical memory consumption.
Materialization was achieved by implementing stored procedures, significantly improving performance while reducing the workload on HANA systems. Blue BI continuously adapted the project design based on test results, ensuring optimal solutions and aiming for excellence.
Deployment
The new architecture was successfully deployed in production, with the presentation layer progressively redirected, enabling the phased decommissioning of Netezza and Dataservices integration flows.
This project’s success established an architecture based on SAP HANA structures, providing a foundation for further analytical implementations.
Why Blue BI
The project’s success hinged on Blue BI’s ability to add value and quality by balancing diverse client needs. The phased migration approach, supported by Blue BI’s expertise, was well-received, delivering positive results at each step.
By adopting tailored strategies and providing technical and implementation excellence, Blue BI achieved seamless migration and performance improvements, enhancing the client’s analytics capabilities while maintaining continuity and ensuring long-term value.
