4 data migrations you need to understand
December 7, 2017
Understand how 4 important types of data migrations work and decide when to enter data manually
Some of our customers use the data management or DIXF functionality that is standard in Microsoft Dynamics 365 for Operations and already was part of Dynamics AX. Issues and limitations in this functionality too often require customizations in the code to support successful migrations. This typically increases project costs and introduces risks. Companies then frequently chose to take advantage of To-Increase Connectivity Studio to keep their risks and costs under control. The solution works so well that enthusiastic customers want to migrate everything. But is that really a good idea? Let’s consider four types of migrations:
- Master data
Master data: Easy if you keep it simple
Master data about, for instance, customers, vendors, products, or trade agreements can be easily imported from a number of sources, including cloud-based or on-premise databases, Excel files based on predefined templates, and CSV files. Best practices makes this task reliable and efficient.
Using Excel spreadsheets, you can load base data and add as many tables as you want. The downside of this approach is that you may end up with hundreds of migrations which have to be designed, maintained, and tested before you perform them.
Transactions: Import the data that matters
You can import transactions in a similar way to master data. However, it is crucial to only import transaction data that is essential for defining a starting point at the moment of going live with your ERP system. Examples of such transactions are:
- Open customer transactions
- Open vendor transactions
- On-hand inventory (including batches and license plates)
- General-ledger balances
In operational data sources you should not import history, such as:
- Inventory transactions
- Ledger transactions
- Invoiced orders
- Finished or ended production orders
Why? The new ERP system will process data differently from the prior system. In addition, the processing of these kinds of data also populates multiple other data sources that have to stay in sync. If historical data is needed for reporting purposes, you should import it in additional tables, so that business analytics solutions can access it.
Companies: Use a template
When you need to migrate multiple companies, you can use a template company to simplify the setup of new companies. Based on the type of company, it should get the parameters, main accounts, and other setup details from the template company. When you use Connectivity Studio, this becomes very easy, as described here.
You can copy company setup data within the same instance of the ERP system or by using ODBC. Doing so requires some upfront effort but can save significant time when you create new companies. The number of companies you have to create determines if the initial investment will pay off.
Setup: Multiple environments make this more efficient
When there is only one single company, it is not very useful to migrate setup data automatically. It’s better to set up the data in the production environment.
In a best-practice approach referred to as DTA(S)P, we recommend the use of several possible environments: development, testing, acceptance, staging under certain conditions, and production. Customizations are developed in the development environment. When the technical unit test (TUT) is executed, they are transferred to the test environment. When the functional unit test (FUT) is executed, they are transferred to the acceptance environment, and so on. Data, on the other hand, usually follows the opposite route. Proven, unchanging master data and its setup are entered or migrated into the production environment and periodically transferred to the other environments.
Proven parameters and an also proven, simple setup can best be entered manually in the production environment. This has several advantages:
- Key users get practice in using the system.
- You don’t need to transfer knowledge and coordinate actions between key users and the employees who configure the migrations.
- The setup is verified and tested after production data is copied to the other environments, and will not be affected anymore.
As a rule of thumb, you should manually enter any tables with fewer than 100 records. Examples might be payment terms, payment modes, delivery terms, delivery modes, company chains, item groups, customer groups, vendor groups, and others.
If you want to learn more about working with Connectivity Studio or the other Business Integration Solutions, please send me a note at firstname.lastname@example.org.