19 Jun

Data Migration for SAP

In most projects test data is obtained from a dedicated data migration team, extracting from legacy systems, transforming to SAP fields, and loading into SAP.

Christian Bergeron (http://christian.bergeron.voila.net/EN.htm) has written a great guide to planning and executing SAP Data Migration, and while most SAP Test Managers won’t need to get their hands dirty in the actual ETL process, it’s important to know and understand the data migration process for a number of reasons:

  • The Data plan usually feeds into one or more of your test phases (for example, Trial Conversion 2 into Integration Testing, Trial Conversion 3 into UAT, etc). It’s critical to ensure that your test phases match up with the data plan and that there is sufficient contingency between the data load planned finish and the test cycle start.
  • Any delay in data load can potentially impact your test schedule. If you are aware of the load sequence you can reschedule tests and reassess test plans so that the impact is mitigated. For example, do you really need those AR/AP Open Items? Can you write on stock or manufacture as part of your testing rather than waiting for the inventory load?
  • On the opposite side of being able to reassess tests and plans is being able to use the load schedule to determine the “must haves” – i.e. what objects are absolute dependencies for being able to start testing.
  • The data migration plan should include details of the environments used and any cutover / client copy / data copy process to your test environment. Will this include data validation and defect fixing before the copy? What is the expected level of quality of data that will hit your test system? Understanding the Data plan and their mitigations will assist your test planning and risk management.
  • Part of your test prep will include identifying the key data required for your test scenarios. By understanding the data migration process and timelines you can better direct your testers to complete test data verification and enable you to plan accordingly.
  • How are the Data team communicating progress in their loads? Ensure you’re on any daily distribution of load status so you can see any delays and plan to prepare accordingly.

And perhaps most importantly:

  • Both Data and Testing can expect to have some crossover in critical business resources – the same people that you want for UAT will be the same people that will be completing data validation (and will be the same people that need to go on training courses or become super users / floorwalkers or write business continuity plans or…. ). By understanding the Data Migration plan you can work with both data and the business to identify those key people early and help to create resource plans to cover those critical phases where they are being pulled from pillar to post.

I’ve attached Christian’s document below (unchanged, as per copyright on page 2).