1 Introduction
Converting data into Merchandising is a critical step to ensure a successful implementation of the solution. Depending on your overall implementation approach, you may phase implementation of different modules over time - for example, starting with foundation data and then later going live with purchase orders and inventory - or you may implement all available Merchandising modules at once. This document will majorly describe the approach of where all modules are expected to be implemented at once, sometimes referred to as a "big bang" conversion.
However, if you are performing a phased conversion, there are conversion APIs designed to enable the subsequent phase conversion of certain entities in the production environment. See Appendix: Conversion APIs for more information about how to load converted data by effectively appending it to the live production data. Also, see the Phased Data Conversation Reference Paper, which discusses the approaches available for Merchandising cloud services phased rollout.
At a high level, data conversions go through the following steps:
The first three steps are expected to be performed in your legacy solutions. The last step will be to load the cleansed, extracted, and transformed data into Merchandising.
Most data entities from legacy applications to Merchandising can be converted through the Data Conversion application. Data files from your legacy solutions can be loaded into this application. After that, data will go through integrity checks and business validations and validated data will be loaded to Merchandising tables. This is a screen-based module to guide you through the conversion process. All data errors are reported at a detailed level to assist you to iteratively correct and re-run through the conversion process. The final converted data will then be exported to production environment.
The Data Conversion application is installed in non-production environments as a part of the Merchandising environment provisioning process in the same database as Merchandising, but in a separate schema.
You are recommended to start analyzing your legacy data along with or before the provisioning process to determine how it maps to the Merchandising data model. This includes looking at data cleansing, extraction procedures, and transformation logic. To support this process, two tools are provided as part of this application - Data Templates and Offline File Validator. See more at Appendix: Offline File Validator. Data templates provide detail on the structure of the data files, whereas the Offline Validator tool allows you to start pre-validating your transformed data. These are described in more detail below.
Once data is extracted, transformed, and templates are populated, that data can be transferred to the conversion environment and loaded into the Data Conversion tool to be validated. The last step, once all conversion related validations and testing is complete, is to load the data into production.
