Dual-write FAQ

This article lists frequently asked questions about dual-write and provides brief answers to help you quickly get the information that you require.

Dual-write setup

Do you plan to enable dual-write to use Dataverse as a hub between multiple finance and operations environments? If Dataverse is used as a hub, data can be synced between two or more finance and operations environments.

The current plan of record is to restrict dual-write to a one-to-one (1:1) mapping between a single finance and operations environment and a single Dataverse environment.

Can I control the sequencing of maps in dual-write, as I can in Data integrator?

Dual-write is transaction-based. For example, if a change in a finance and operations app triggers synchronization of multiple maps with Dataverse, by default, those changes will be sequenced in the order in which they are updated in the database. This pattern makes more sense in the context of initial synchronization. The system provides related table maps in a specified order, and you can reorder the list so that it best suits your environment.

Do application users require any special permissions to enable or configure dual-write?

You must have two Azure Active Directory (Azure AD) applications set up for the finance and operations environment and two application users set up in the Dataverse environment. These application users should contain the appropriate application IDs. For the connection to work properly, you must give the applications the relevant table permissions by using a security role. For more information, see Verify requirements and grant access.

Do end users require any special permissions to enable or configure dual-write?

End users who are configuring dual-write mappings should have System Administrator security roles assigned in both Dataverse and finance and operations environments.

Dual-write mappings can be accessed by multiple users, as long as all the users and environments belong to a single tenant, and the user has the required security and licenses assignment.

I have multiple legal entities. Some of my maps are legal table–specific or valid for only some of the legal entities. What is the best way to address this requirement? Can I apply a filter such as Company = USMF to address it?

Legal table mapping can be done when the Dataverse environment is linked. You can't map table maps to a specific legal entity.

If dual-write solutions are installed in Dataverse, can I uninstall them?

Dual-write solutions are managed solutions that can be uninstalled. However, when a managed solution is uninstalled, all components in the solution are deleted. Any data that is stored in the components is also deleted. For more information, see Maintain managed solutions.

I have data in both a customer engagement app and a finance and operations app, and I bootstrap my existing data in the customer engagement app. If my data isn't currently aligned, can I specify a master source for the initialization run, so that all differences are applied to the target?

After the bootstrapping is done, you can configure the initial synchronization to apply differences and select a master. For more information about bootstrapping, see Bootstrap with company data FAQ. For more information about the initial synchronization, see Enable table maps for dual-write.

Dual-write administration and management

What is the purpose of the integration key, and is it mandatory?

The integration key is the natural key that uniquely identifies rows. Integration keys are required only for Dataverse tables. You can manually create an integration key in dual-write. An integration key can also be automatically created from the table's alternate keys, if an alternate key is already provided for the table. Integration keys are used for the same purpose as alternate keys: they provide an efficient and accurate way to integrate data with external systems. Integration keys are essential in cases where an external system doesn't store the globally unique identifiers (GUIDs) that uniquely identify rows in Dataverse.

Dual-write uses integration keys to uniquely identify rows, by using one or more table column values that represent a unique combination. For example, to identify an account row by using an integration key, you can use the account number column. Alternatively, you can use the account number column together with other columns that have values that should not change. For more information, see Define alternate keys using Power Apps portal.

It's important that keys be matched between the finance and operations environment and the Dataverse environment. Otherwise, issues might occur during the initial synchronization phase.

How do I move table maps between environments? Is version control supported for table maps?

You can export maps and then import them into a different environment. You can automate the process by using Azure DevOps. You can have version control on your dual-write mappings, because the mappings are solution-aware components. For more information, see Update table maps and export them to other environments as a solution.

Where can I find examples and patterns for filtering dual-write maps?

For basic filtering examples, see Filter your data.

For more advanced examples for Dataverse, see Filter results. Nested lookup isn't supported in dual-write source filters. Only standard filter operators directly against table columns are supported.

For more advanced finance and operations filters, see Using Expressions in Query Ranges and Advanced filtering and query syntax.

Dual-write live synchronization introduces tight coupling across applications. What happens if one side fails? Will the other side fail too?

When the integration is in live sync mode, if the sync fails on one of the apps, then the other app will fail as well and users will receive an error. When the integration is paused, changes are staged. They are then written when the target system is up and running. For more information about how to automatically pause integrations, see Alert notifications

When live synchronization is paused and then resumed, does it follow the sequence of changes? For example, if the Name column in the finance and operations app is changed from NameA to NameB to NameC, is customer engagement data changed from NameA to NameB to NameC, or is it changed directly from NameA to NameC?

The integration follows the complete sequence of changes. In the example, the customer engagement app data is changed from NameA to NameB to NameC.

How do I handle a finance and operations database transfer from PROD to STAGE? What is the effect on dual-write? After the transfer, the systems are no longer in sync. Is the synchronization done automatically?

Each linked environment-pair (finance and operations apps environment and Dataverse environment) should be treated as a single unit and refreshed accordingly. For example, if you are refreshing a sandbox from production, then both finance and operations app sandbox environment and the Dataverse sandbox environment should be refreshed from their production counterparts. If dual-write is already used in target environments, those environments need to be unlinked. After the data refresh on target environments, these tables should be cleaned up:

• finance and operations apps tables: DualWriteProjectConfiguration, DualWriteProjectFieldConfiguration, and BusinessEventsDefinition.
• Dataverse tables: DualwriteRuntimeConfiguration.

The environments need to be relinked and maps reactivated manually.

I need real-time integration, and I want to move some tables or scenarios from Data integrator to dual-write. How do I migrate, and what are the implications of changing my integration pattern?

For information about how to migrate Prospect to cash to dual-write, see Migrating data from Data Integrator to Dual Write. In general, three things might change during migration:

• Manual migration of the maps from Data integrator to dual-write
• Table changes, because of the absence of advanced query capabilities
• Data migration, because of adaptation to new concepts such as company striping

On finance and operations data tables, can I develop unbounded columns that flow to Dataverse by using dual-write?

Yes. You can use both computed columns and virtual columns. However, you should monitor the performance overhead from the additional X++ logic that is required for reads and writes. Round-tripping within the same transaction isn't allowed. Therefore, you should avoid using virtual columns to transform or calculate additional values through X++ and expect that to go back to Dataverse within the same transaction.

When I use the Dataverse offline app, what happens if I can't sync the data after reconnection? Does this situation cause an inconsistent state between the Dataverse environment and the finance and operations environment?

You can interact with Dataverse data offline when using the Dynamics 365 for phones app or the Field Service Mobile app in offline mode. In both apps, data is stored offline and can be synced with the server at your discretion. If there are errors when the offline data is synced with the server, and updates can't be done because the other environment is failing, data sync will fail, and Dataverse will not be updated. When the integration is paused, you can re-run the sync and save your updates on the server. These changes will be staged and then synced with the finance and operations environment when the mapping is up and running again. For more information, see Run model-driven apps and canvas apps on Power Apps mobile.

Mapping concepts between apps

How are number sequences handled? For example, the customer account number is automatically generated in finance and operations apps, but it's added manually in customer engagement apps.

Number sequences for finance and operations apps and customer engagement apps aren't connected. In a scenario that involves a multi-mastered table, you must either plan for separate number sequence formats or create a range for each app. Here are some examples:

• In the finance and operations app, use F0001, F0002, F0003. In the customer engagement app, use C0001, C0002, C0003.
• In the finance and operations app, use US0001 to US4999. In the customer engagement app, use US5000 to US9999.

If a table is created in only one system, set up the number sequence in the source app only. For more information, see Autonumber columns.

Can I map a company-specific table in a customer engagement app with a global table in a finance and operations app, or a global table in a customer engagement app with a company-specific table in a finance and operations app?

Dual-write supports mappings only between cross-company tables or company-specific tables from both sides.

How do I make a company-specific table in Dataverse?

You can make Dataverse custom tables company-specific by adding a many-to-one (N:1) relationship between your custom tables and the out-of-box company table. You should also include the company foreign key as part of the table key. For more information, see Company concept in Dataverse.

To enable table maps for dual-write, you must define an alternate key in Dataverse. The value of the alternative key in Dataverse must match the key that is defined in the finance and operations app. For more information, see Criteria for linking tables.

Can I merge accounts in customer engagement apps and party records in finance and operations apps while using dual-write?

No, there's no parity between the merging functionalities in finance and operations apps and customer engagement apps. As a result, when a dual-write mapping is present on a table:

• Merging accounts in customer engagement apps won't execute.
• Merging party records in finance and operations apps may result in data mismatch.

Is there a document about best practices for table usage? Should I use Customers V2, Customers V3, or Customer Details? What is the difference between these tables, and what is the use case for each?

You should use the out-of-box scenarios if you can, because they cover common scenarios such as customer/vendor integration.