CertSafari

    Free Microsoft Dynamics 365: Finance and Operations Apps Developer (MB-500) - December 18 onwards Sample Questions

    35 free sample questions from our bank of 350+, covering every exam domain, with answers and detailed explanations. Updated November 2026.

    1: Plan the architecture and solution design

    1.1: Plan for the major components of finance and operations apps

    1.A developer who wants AI chat assistance for writing X++ inside the IDE should use the ______ control in Visual Studio 2022.

    1. A.Task recorder
    2. B.GitHub Copilot Chat
    3. C.Trace Parser
    Show answer & explanation

    Correct answer: B — GitHub Copilot Chat

    • A. Incorrect. Task recorder captures business process steps and does not generate code.
    • B. Correct. GitHub Copilot Chat in Visual Studio 2022 provides AI assistance for X++ development.
    • C. Incorrect. Trace Parser analyses execution traces and offers no generative assistance.

    1.1: Plan for the major components of finance and operations apps

    2.A partner wants a Copilot Studio agent to look up customer balances and run business logic in a finance and operations environment without writing a custom connector or custom service per entity. Which capability should the partner use?

    1. A.Custom X++ service classes in finance and operations, with the agent calling one JSON endpoint per entity
    2. B.A batch job that exports balances to Azure Blob Storage hourly, with the agent querying the exported files
    3. C.The Dynamics 365 ERP MCP server, with the agent invoking its tools to read and write data and run logic
    4. D.Excel add-in publishing of each data entity, with the agent reading the refreshed workbook from SharePoint
    Show answer & explanation

    Correct answer: C — The Dynamics 365 ERP MCP server, with the agent invoking its tools to read and write data and run logic

    • A. Incorrect. Custom X++ services can expose data and logic, but each entity needs its own service class and endpoint to build and maintain, which is the per-entity custom work the partner wants to avoid.
    • B. Incorrect. Hourly exports leave the agent with stale files and no ability to run business logic in the ERP.
    • C. Correct. The Dynamics 365 ERP MCP server gives agents ready-made tools for finance and operations data operations and business logic, so no custom connector or per-entity service has to be written.
    • D. Incorrect. Publishing entities to Excel is a manual office-integration feature and is not a way for an agent to call business logic.

    1.2: Implement application lifecycle management (ALM)

    3.UAT sign-off just finished for a deployable package in a finance and operations apps project. The team wants it in production. Which approach matches the supported production servicing process?

    1. A.Mark the package as a release candidate in the LCS asset library, then submit a service request for Service Engineering to schedule it
    2. B.Add a production stage to the Azure DevOps release pipeline, then approve the stage so the pipeline deploys the package
    3. C.Run the package installer on the production AOS machines from a remote session the partner opens, then restart the AOS service
    4. D.Import the package through the Data management workspace in production, then run a database synchronization from the developer tools
    Show answer & explanation

    Correct answer: A — Mark the package as a release candidate in the LCS asset library, then submit a service request for Service Engineering to schedule it

    • A. Production deployment is self-service but manual. The package is flagged as a release candidate in the LCS asset library, and a service request lets Service Engineering schedule the deployment.
    • B. Pipeline-based automated deployment from Azure DevOps covers non-production environments only. Production is serviced by Microsoft, so no pipeline stage can apply a package there, even with an approval.
    • C. Production AOS machines are managed by Microsoft, and neither customers nor partners get remote access to them. A hand-run installer is therefore unavailable, and the partner cannot open such a session.
    • D. The Data management workspace imports data entities, not deployable code packages. A database synchronization from developer tools does not move code into production either.

    1.2: Implement application lifecycle management (ALM)

    4.Two developers regularly deploy to the same shared unified developer environment from different machines, and the environment starts behaving inconsistently after one developer switches Git branches. What practice prevents this?

    1. A.Keep the environment in administration mode between deployments, letting only one developer connection deploy at a time
    2. B.Run a full database synchronization before each deployment, rebuilding the schema from the metadata of the last machine
    3. C.Store custom metadata in Azure DevOps version control, giving a referenced record of what is currently deployed
    4. D.Grant developers the System Administrator role in Dataverse, letting deployments overwrite earlier customizations
    Show answer & explanation

    Correct answer: C — Store custom metadata in Azure DevOps version control, giving a referenced record of what is currently deployed

    • A. Administration mode restricts access to admin users for maintenance. Limiting connections does not make the metadata on different machines and branches agree, because each deployment still pushes whatever that machine holds.
    • B. Database synchronization only aligns the schema with metadata that is already deployed. It cannot decide whose code is correct, so rebuilding from the last machine's metadata carries that machine's drift forward.
    • C. Version control gives one referenced history of the custom metadata. Each deployment is then built from tracked contents and can be checked against what is already deployed, whichever machine or branch it comes from.
    • D. A security role controls who may deploy, not whether the deployed content matches what is already there. Letting deployments overwrite earlier ones just replaces one machine's state with another's.

    2: Apply developer tools

    2.1: Customize finance and operations apps by using Visual Studio

    5.The read-only ____ property on an element in Application Explorer is used for synchronization and upgrades.

    1. A.LegacyID
    2. B.ConfigurationKey
    3. C.Origin
    Show answer & explanation

    Correct answer: C — Origin

    • A. Incorrect: LegacyID holds the identifier kept from earlier versions and is not the GUID used for synchronization.
    • B. Incorrect: ConfigurationKey determines whether an element is visible to users, based on enabled configuration.
    • C. Correct: Origin holds a GUID that identifies the element for synchronization and upgrade processes.

    2.2: Manage source code and artifacts by using Microsoft Azure DevOps and GitHub version control

    6.A developer's local `main` is behind the remote. To download the new commits and integrate them into the current branch in a single step, the developer should run `git ___`.

    1. A.fetch
    2. B.stash
    3. C.pull
    Show answer & explanation

    Correct answer: C — pull

    • A. Fetch downloads remote commits but leaves the current branch unchanged, so integration would still be needed.
    • B. Stash shelves uncommitted local edits temporarily and does not retrieve anything from the remote.
    • C. Pull performs a fetch and then merges (or rebases) the remote changes into the current branch.

    2.1: Customize finance and operations apps by using Visual Studio

    7.A developer created a new class through Add > New Item, but later finds it was stored in the Contoso.Base model instead of the intended Contoso.Sales model. What most likely caused this?

    1. A.The class name started with the prefix Base, and Visual Studio saves the element in the model whose name matches that prefix.
    2. B.Contoso.Base was the first model in the same package folder, and Visual Studio saves new elements in the first model it finds.
    3. C.The Origin property on the class held a GUID from Contoso.Base, and Visual Studio saves the element in the model owning that GUID.
    4. D.The project's Model property was set to Contoso.Base, and Visual Studio saves new elements in the model the project targets.
    Show answer & explanation

    Correct answer: D — The project's Model property was set to Contoso.Base, and Visual Studio saves new elements in the model the project targets.

    • A. Incorrect: Visual Studio does not match a class name prefix to a model name. Prefixes are only a naming convention, so Base in the name would not move the element into Contoso.Base.
    • B. Incorrect: Visual Studio does not pick the first model it finds in a package folder. The target model comes from the project the item is added to, not from the order of models on disk.
    • C. Incorrect: Origin is a GUID the system maintains for synchronization and upgrade tracking. It does not decide which model owns an element, so a Contoso.Base GUID would not redirect the class.
    • D. Correct: each Dynamics 365 project is tied to one model, and elements added through Add > New Item are saved in that model. A project pointing at Contoso.Base therefore stores the class there.

    2.2: Manage source code and artifacts by using Microsoft Azure DevOps and GitHub version control

    8.After running a pull in Visual Studio, a developer sees a merge conflict in the AxClass XML file for `SalesInvoiceHelper` because a teammate edited the same method. The developer wants to keep both people's changes. What should they do?

    1. A.Commit the local version of the class, then run `git push --force-with-lease` so the remote branch carries the local method body.
    2. B.Abort with `git merge --abort`, then pull again using `-X theirs` so Git settles the overlapping hunks in the AxClass XML.
    3. C.Run `git checkout --ours` on the AxClass XML, stage the file, and finish the merge so the local version of the method is kept.
    4. D.Open the merge editor, edit the two versions of the method into one body, stage the file, commit the merge, and rebuild the model.
    Show answer & explanation

    Correct answer: D — Open the merge editor, edit the two versions of the method into one body, stage the file, commit the merge, and rebuild the model.

    • A. Committing the local class and pushing it over the remote branch replaces the teammate's method edits instead of combining them with the developer's. The conflicting hunks are never merged.
    • B. Aborting discards the merge in progress. Pulling again with `-X theirs` settles each conflicting hunk in favor of the incoming commits, so the developer's edits to the method are dropped.
    • C. During a merge, `git checkout --ours` takes the local side of the file wholesale. The teammate's changes to the same method are lost, so only one person's edits survive.
    • D. The merge editor lets the developer build a single method body from both sets of edits. Staging marks the file resolved, the merge commit finishes the pull, and the rebuild checks that the combined X++ compiles.

    3: Design and develop AOT elements

    3.2: Create and extend the data model

    9.A custom public data entity is called through OData and fails because it has no usable key. The entity uses `ContosoCode` as the natural identifier of its root table. What should the developer do?

    1. A.Create a unique index over the mapped ContosoCode field on the entity and set it as the entity's primary key
    2. B.Set the public collection name to the plural of the entity name and rebuild so OData derives a key from the name
    3. C.Enable data management capabilities so the generated staging table supplies a key that OData can use directly
    4. D.Change the primary data source to a view so a key is inferred automatically from the view's first column
    Show answer & explanation

    Correct answer: A — Create a unique index over the mapped ContosoCode field on the entity and set it as the entity's primary key

    • A. Correct: a public entity needs a unique key made of mapped fields so OData can address each record.
    • B. Incorrect: the collection name only names the resource set; it does not provide a key.
    • C. Incorrect: the staging table serves asynchronous data management jobs and does not define the OData entity key.
    • D. Incorrect: entities never infer keys from the data source, so the key must be modelled explicitly regardless of source type.

    3.1: Create and extend UI elements

    10.A developer must block saving a sales line when a custom field is blank, but only when the line is edited on the SalesTable form, not during imports or other code paths. Which extension fits?

    1. A.Wrap `validateWrite` in a final class decorated with `[ExtensionOf(formDataSourceStr(SalesTable, SalesLine))]`, calling `next` and returning false for a blank field.
    2. B.Wrap `validateWrite` in a final class decorated with `[ExtensionOf(formStr(SalesTable))]`, calling `next` and returning false for a blank field.
    3. C.Wrap `validateWrite` in a final class decorated with `[ExtensionOf(tableStr(SalesLine))]`, calling `next` and returning false for a blank field.
    4. D.Register a `[FormDataSourceEventHandler]` method for the `ValidatedWrite` event of the SalesLine data source, returning false for a blank field.
    Show answer & explanation

    Correct answer: A — Wrap `validateWrite` in a final class decorated with `[ExtensionOf(formDataSourceStr(SalesTable, SalesLine))]`, calling `next` and returning false for a blank field.

    • A. A Chain of Command extension on the SalesLine data source of the SalesTable form wraps the data source's own `validateWrite`. It runs only when a line is saved on that form, so imports, data entities and other X++ code that write the table directly are not affected.
    • B. `validateWrite` for the SalesLine data source belongs to the data source, not to the SalesTable form class. A form-level extension has no such method to wrap, so the check would never run when a line is saved.
    • C. A wrapper on the SalesLine table runs for every caller that writes the table, including imports, data entities and X++ code. The blank-field check would then also block those paths, which must stay unaffected.
    • D. The handler for the data source's `ValidatedWrite` event is a void method that receives event arguments. It has no return value to set, so returning false from it cannot veto the save, even though the event is scoped to the form data source.

    3.1: Create and extend UI elements

    11.A new list page uses the Simple List pattern and must stay compliant with it. Visual Studio reports pattern validation errors because a Tab control sits directly under the design's custom container. What resolves the errors?

    1. A.Set the Tab control's `Visible` property to No, which makes the validator skip the control when it checks the design.
    2. B.Add a pattern extension element to the model, which whitelists the Tab control as a child of the custom container.
    3. C.Change the form's Style property to Dialog, which applies the dialog layout rules to the list page during validation.
    4. D.Restructure the design's containers as action pane, custom filter group, then grid, which is the order the validator expects.
    Show answer & explanation

    Correct answer: D — Restructure the design's containers as action pane, custom filter group, then grid, which is the order the validator expects.

    • A. Pattern validation examines which controls sit under which containers, not their runtime visibility. A hidden Tab control is still in the wrong place, so the errors remain.
    • B. Patterns are defined by the platform, and adding an extension element to the model does not loosen their structural rules or permit the Tab control under that container.
    • C. The Dialog style applies a different pattern with its own layout rules. That abandons Simple List, which the new list page must remain compliant with, and changes how the page behaves.
    • D. Pattern validation checks that the design's containers follow the required structure. Simple List expects an action pane, a custom filter group and a grid, so restructuring to that order clears the errors.

    3.2: Create and extend the data model

    12.A developer is documenting how AOT data-model elements differ. Which three statements are accurate?(Select 3)

    1. A.A view is synchronized as a SQL view object in the database and can be queried but not written to
    2. B.A map defines a shared set of fields over several tables and has no table or view in SQL
    3. C.A query is AOT metadata that X++ runs through QueryRun and has no database object of its own
    4. D.A map is synchronized as a physical table and stores the rows of the several tables it covers
    5. E.A data entity is exposed as a read-only SQL view and imports must write to the tables directly
    Show answer & explanation

    Correct answers: A, B, C — A view is synchronized as a SQL view object in the database and can be queried but not written to; A map defines a shared set of fields over several tables and has no table or view in SQL; A query is AOT metadata that X++ runs through QueryRun and has no database object of its own

    • A. Correct: a view becomes a SQL view object when the database is synchronized, so it can be read through SQL but cannot be written to.
    • B. Correct: a map is only metadata that lets code treat fields of different tables as one shape, so synchronization creates no table or view for it.
    • C. Correct: a query is AOT metadata describing data sources, joins and ranges. It is executed by X++ through QueryRun and creates no database object itself.
    • D. Incorrect: a map is not synchronized to the database and holds no rows. The data stays in the underlying tables it maps.
    • E. Incorrect: data entities support create, update and delete, and they are the main route for data import and export, so imports do not have to bypass them.

    4: Develop and test code

    4.2: Develop object-oriented code and implement code extensions

    13.A developer needs a method in a base class to be unavailable for overriding in any subclass. The method must be declared ____.

    1. A.abstract
    2. B.protected
    3. C.final
    Show answer & explanation

    Correct answer: C — final

    • A. Incorrect: abstract methods have no body and subclasses are required to override them, which is the opposite of the goal.
    • B. Incorrect: protected only limits visibility to the class and its subclasses; derived classes may still override a protected method.
    • C. Correct: the final modifier prevents any subclass from overriding the method.

    4.3: Implement finance and operations app framework functionality

    14.A packed query in a nightly batch job must always filter on the user who runs the job at execution time. The range value should be set with the expression `(______())`, evaluated by `SysQueryRangeUtil` when the query runs.

    1. A.`currentUserId`
    2. B.`curUserId`
    3. C.`currentUserName`
    Show answer & explanation

    Correct answer: A — `currentUserId`

    • A. Correct. SysQueryRangeUtil exposes currentUserId, and a range value expression is resolved when the query runs, so it matches the executing user.
    • B. Incorrect. curUserId is a global X++ function that would be evaluated when the query is built, so a packed query would keep the user who built it.
    • C. Incorrect. currentUserName is not a SysQueryRangeUtil method and would not match the user id values stored in the field.

    4.1: Develop X++ code

    15.A report job uses `while select * from InventTrans where InventTrans.DatePhysical >= startDate` over about 2 million rows to total `CostAmountPosted`, but never reads other fields. Which two changes most improve performance?(Select 2)

    1. A.Aggregate with `select sum(CostAmountPosted) from inventTrans where ...`, so the server returns a single row
    2. B.Add `forceliterals` to the select, so the optimizer plans around the literal date rather than a parameter
    3. C.Apply `forupdate` to the select, so the server holds a cursor and streams the rows through the loop faster
    4. D.Select only the field with `while select CostAmountPosted from inventTrans`, so each fetched row is narrower
    5. E.Use `crosscompany` on the select, so the scan spans legal entities and the optimizer parallelizes it
    6. F.Add `firstonly` to the select, so the server stops scanning after the first row matching the date filter
    Show answer & explanation

    Correct answers: A, D — Aggregate with `select sum(CostAmountPosted) from inventTrans where ...`, so the server returns a single row; Select only the field with `while select CostAmountPosted from inventTrans`, so each fetched row is narrower

    • A. Correct. An aggregate select makes the database compute the sum and return one row, which removes the transfer of about 2 million rows and the loop that adds them up.
    • B. Incorrect. forceliterals embeds the values in the SQL text, which disables parameterization, bloats the plan cache and exposes the statement to injection. It does nothing about the volume of rows and columns transferred, which is the actual cost here.
    • C. Incorrect. forupdate makes the server take update locks and fetch rows in a heavier form. It does not speed up reading, and the job only reads values.
    • D. Correct. A field list means only CostAmountPosted is fetched for each row instead of the whole wide InventTrans record, which cuts memory and network cost when row values are still needed.
    • E. Incorrect. crosscompany widens the query to other legal entities, so it reads more data and could change the total. It does not make the scan run in parallel.
    • F. Incorrect. firstonly returns a single record, so the loop would see one row and the total would be wrong. It does not make the correct total faster.

    4.2: Develop object-oriented code and implement code extensions

    16.A team writes a Chain of Command wrapper for a public instance method with several parameters. Which statements about the wrapper are correct? (Select all that apply.)(Select 3)

    1. A.The wrapper method must keep the same name, parameter list, and return type as the method it wraps.
    2. B.The wrapper can read protected members of the augmented class directly, with no accessor methods needed.
    3. C.The wrapper can wrap private methods of the augmented class, provided the signature matches exactly.
    4. D.The wrapper class must carry `[SubscribesTo]` for the augmented class to be picked up at startup.
    5. E.The wrapper must call `next` as its final statement, after its own logic has finished running.
    6. F.The wrapper class must be declared `final` and use `[ExtensionOf]` to name the class it augments.
    Show answer & explanation

    Correct answers: A, B, F — The wrapper method must keep the same name, parameter list, and return type as the method it wraps.; The wrapper can read protected members of the augmented class directly, with no accessor methods needed.; The wrapper class must be declared `final` and use `[ExtensionOf]` to name the class it augments.

    • A. Correct: a Chain of Command wrapper mirrors the signature of the method it wraps, and the call to `next` forwards the same arguments to the base method.
    • B. Correct: a Chain of Command extension class can read protected members of the augmented class directly, so no accessor methods are needed.
    • C. Incorrect: matching the signature does not make a private method wrappable. Only public and protected methods can be wrapped.
    • D. Incorrect: `[SubscribesTo]` is for event handlers that subscribe to delegates. A Chain of Command class is tied to its target by `[ExtensionOf]` alone.
    • E. Incorrect: `next` can be called before, after, or around the wrapper's own logic. Nothing requires it to be the last statement.
    • F. Correct: the extension class must be declared `final` and use `[ExtensionOf]` to name the class being augmented.

    4.4: Perform testing

    17.An architect wants each test class to declare which production class it validates so that the relationship is explicit in metadata. Which attribute does the SysTest framework provide for this?

    1. A.SysTestCheckInTestAttribute, which tags the test class as a check-in test so it runs with changes to the code it covers.
    2. B.SysTestTargetAttribute, which names the class, table or other element that the test class is intended to cover.
    3. C.SysTestMethodAttribute, which marks a test method and takes the name of the production class as its argument.
    4. D.SysTestInactiveTestAttribute, which marks the test class as inactive so it stays registered against its production class.
    Show answer & explanation

    Correct answer: B — SysTestTargetAttribute, which names the class, table or other element that the test class is intended to cover.

    • A. Incorrect: SysTestCheckInTestAttribute only classifies a test as one to run at check-in. It takes no production class, so the relationship is not declared in metadata.
    • B. Correct: SysTestTargetAttribute declares the element under test, making the test-to-code relationship explicit.
    • C. Incorrect: SysTestMethodAttribute marks a method as a test and has no argument for a production class. It cannot declare which class a test class validates.
    • D. Incorrect: SysTestInactiveTestAttribute disables a test so it is skipped. It does not link the test to a production class, so it cannot make that relationship explicit.

    4.4: Perform testing

    18.A developer opens Test Explorer to validate a change. Which two actions can be taken there? (Select all that apply.)(Select 2)

    1. A.Build a deployable package from the model under test
    2. B.Capture a failed test run as a Task recorder recording
    3. C.Run a chosen subset of the tests from the test list
    4. D.Debug a selected test with breakpoints set in X++ code
    5. E.Edit a test method's X++ source in the results pane
    Show answer & explanation

    Correct answers: C, D — Run a chosen subset of the tests from the test list; Debug a selected test with breakpoints set in X++ code

    • A. Incorrect: Test Explorer only lists, runs and debugs tests. Deployable packages are created through the Dynamics 365 build and deployment tooling and released through Lifecycle Services, so a developer validating a change cannot do this there.
    • B. Incorrect: Test Explorer has no feature that turns a test run into a recording. Task recorder captures a user's actions in the client web interface, which is a separate tool from the test results.
    • C. Correct: Test Explorer can run the whole set of tests or only the ones the developer selects, which keeps feedback quick while validating a change.
    • D. Correct: a selected test can be started in debug mode, so breakpoints set in the X++ code are hit while the test executes.
    • E. Incorrect: the results pane shows test names and outcomes and is not an editor. Changes to a test method's X++ source are made in the code editor, not in Test Explorer.

    4.3: Implement finance and operations app framework functionality

    19.A team converts a RunBase class that processes vendor invoices to the SysOperation framework. Which THREE statements describe pieces they must build?(Select 3)

    1. A.The data contract is a class marked with `DataContractAttribute` whose parameters are exposed through parm methods marked with `DataMemberAttribute`.
    2. B.The service class exposes the business logic in a method marked with `SysEntryPointAttribute`, which the framework invokes with the contract.
    3. C.The controller extends `SysOperationServiceController`, names the service class and method, and calls `startOperation()` to launch the process.
    4. D.The contract class implements `pack()` and `unpack()` with a `#CurrentList` macro to persist dialog values between consecutive runs.
    5. E.The service class extends `RunBaseBatch`, which lets the controller schedule its operation method on a batch server group of choice.
    Show answer & explanation

    Correct answers: A, B, C — The data contract is a class marked with `DataContractAttribute` whose parameters are exposed through parm methods marked with `DataMemberAttribute`.; The service class exposes the business logic in a method marked with `SysEntryPointAttribute`, which the framework invokes with the contract.; The controller extends `SysOperationServiceController`, names the service class and method, and calls `startOperation()` to launch the process.

    • A. Correct. The contract class carries the parameters; its members are exposed through DataMemberAttribute parm methods and become dialog fields.
    • B. Correct. The service method with SysEntryPointAttribute holds the business logic and receives the contract as its argument.
    • C. Correct. The controller ties the contract, the service class and the method name together and starts the operation with startOperation().
    • D. Incorrect. The framework serializes the contract through its data contract attributes, so a hand-written pack/unpack with a #CurrentList macro is a RunBase pattern that the converted class does not need to build.
    • E. Incorrect. A service class is a plain class and does not derive from RunBaseBatch. Batch execution is chosen through the controller's execution mode, not through the service's base class.

    5: Implement reporting

    5.1: Select reporting tools in finance and operations apps

    20.A workspace Analytics tab shows an embedded report that a developer added in Visual Studio as a model resource. The report file stored in that resource is a ____ file.

    1. A..pbix
    2. B..rdl
    3. C..rptdesign
    Show answer & explanation

    Correct answer: A — .pbix

    • A. Correct. Embedded Power BI reports are authored in Power BI Desktop and added as .pbix resources to the model.
    • B. Incorrect. The .rdl format is used by SSRS reports, which render on the report server and not in an Analytics tab.
    • C. Incorrect. The .rptdesign extension belongs to BIRT reports and is not used by finance and operations apps.

    5.2: Design, create, and revise Dynamics 365 reports

    21.A multi-page SSRS ledger report must show the balance carried forward from the previous page in each page header. What is the supported way to implement it in the report design?

    1. A.A report controller class that sets ranges and selects the design before executing the report
    2. B.A UI builder class that hides the dialog fields and fills their values from the session context
    3. C.Add a report data provider that filters the rows by legal entity and chooses the design inside processReport
    4. D.Create a security privilege on the report menu item so only the current legal entity's users may open the report
    Show answer & explanation

    Correct answer: B — A UI builder class that hides the dialog fields and fills their values from the session context

    • A. The provider cannot know where page breaks fall at render time, since pagination happens in the report renderer. A per-page row cannot be computed in advance.
    • B. The Precision Designer supports embedded VB code in the report's Code section. Functions there track running totals so header and footer expressions can reference previous-page balances.
    • C. A data provider only fills the dataset. The design is chosen by the controller before rendering begins, not inside processReport.
    • D. Security privileges control who can open a menu item. They do not apply session-based ranges or select among designs.

    5.1: Select reporting tools in finance and operations apps

    22.Finance leaders want interactive analytical tiles in a workspace Analytics tab that use pre-aggregated measures and do not load the transactional database. Which datastore backs this scenario?

    1. A.The entity store, an Azure SQL analytical database refreshed on a schedule from aggregate measurements and read by Power BI reports embedded in the workspace
    2. B.A bring-your-own-database instance, a customer Azure SQL database loaded on a schedule with entity exports and read by Power BI reports embedded in the workspace
    3. C.The AOS transactional database, an Azure SQL database holding the live tables that users post to, read by Power BI reports embedded in the workspace via DirectQuery
    4. D.The report server catalog, a SQL Server database filled with aggregated rows by SSRS dataset providers and read by Power BI reports embedded in the workspace
    Show answer & explanation

    Correct answer: A — The entity store, an Azure SQL analytical database refreshed on a schedule from aggregate measurements and read by Power BI reports embedded in the workspace

    • A. Correct. The entity store is the analytical database that holds aggregate measurements and is refreshed on a schedule. Embedded Power BI reads it, so analytical queries stay off the transactional database.
    • B. Incorrect. A bring-your-own-database instance is a customer-managed export target holding entity data. It is not the store of aggregate measurements that the embedded workspace tiles rely on.
    • C. Incorrect. DirectQuery against the live transactional tables puts analytical query load on the database users post to, which is exactly what the finance leaders want to avoid.
    • D. Incorrect. The report server catalog stores report definitions and metadata. SSRS dataset providers do not write aggregated rows into it, so it is not a source for Power BI measures.

    5.2: Design, create, and revise Dynamics 365 reports

    23.A customer is evaluating SSRS for compliance-oriented purchase order documents. Which TWO built-in capabilities of SSRS-based reports in finance and operations apps support that choice?(Select 2)

    1. A.Batch framework scheduling that runs the report unattended and emails it to recipients
    2. B.Print archive that stores a copy of the generated document for later retrieval
    3. C.Excel integration that builds a pivot table from the report data on the user's desktop
    4. D.Workspace embedding that renders the report as a native chart, configured without code
    5. E.Power BI publishing that exposes the report output as a dataset for external subscribers
    Show answer & explanation

    Correct answers: A, B — Batch framework scheduling that runs the report unattended and emails it to recipients; Print archive that stores a copy of the generated document for later retrieval

    • A. SSRS reports can be run on a schedule through the batch framework and sent by email. This gives recurring, hands-off delivery of purchase order documents.
    • B. Print archive stores a copy of a generated business document so it can be retrieved later. That retained record supports audit and regulatory needs for purchase order documents.
    • C. Pivot tables are built by users inside Excel for their own analysis. SSRS does not generate them on a desktop, and they add no retention or delivery control for formal documents.
    • D. Workspace charts are embedded controls that developers model against data sources. An SSRS document report is not converted into one, and a chart does not preserve a purchase order document.
    • E. SSRS reports are not published as Power BI datasets. Power BI content is authored separately against aggregate data, and it does not archive or distribute formal purchase order documents.

    6: Integrate and manage data solutions

    6.2: Implement data integration concepts and solutions

    24.A client submits several OData requests in one HTTP call. The request must be sent as a POST to the ___ endpoint with Content-Type multipart/mixed.

    1. A./api/services/$batch
    2. B./data/$metadata
    3. C./data/$batch
    Show answer & explanation

    Correct answer: C — /data/$batch

    • A. The services path invokes custom service operations; no batch route exists there for OData requests.
    • B. The metadata endpoint only returns the entity model document and does not process request bodies.
    • C. The batch endpoint under /data is the single URL that accepts multipart/mixed bodies containing multiple OData requests.

    6.4: Integrate with Microsoft Power Platform, Microsoft 365, and Azure

    25.A dual-write link connects a finance and operations environment with Dataverse as a ______ relationship between the two environments.

    1. A.one-to-one
    2. B.one-to-many
    3. C.many-to-many
    Show answer & explanation

    Correct answer: A — one-to-one

    • A. Correct. A finance and operations environment is linked to exactly one Dataverse environment for dual-write, and vice versa.
    • B. One app environment cannot feed several Dataverse environments through a single dual-write link.
    • C. Many-to-many linking is not supported; the mapping is a single pair.

    6.1: Evaluate data integration patterns and scenarios

    26.A developer must let a Power Automate flow run an existing X++ static method that posts a ledger journal, calling it through the same OData service used for entities. Which change exposes the method?

    1. A.Set IsPublic to Yes on the data entity, which publishes its static methods as callable operations on the OData endpoint
    2. B.Add SysODataActionAttribute to a static method on the entity, which publishes it as an action the flow calls with HTTP POST
    3. C.Register the method in the entity's staging class, which makes the data management framework run it during a package import
    4. D.Define a business event for the method, which lets the flow trigger the journal posting by subscribing to that event
    Show answer & explanation

    Correct answer: B — Add SysODataActionAttribute to a static method on the entity, which publishes it as an action the flow calls with HTTP POST

    • A. IsPublic only publishes the entity set and its fields on the OData endpoint. Static methods on the entity stay unexposed until each one is decorated, so the flow still has no way to call the posting method.
    • B. SysODataActionAttribute on a static entity method publishes it as a bound OData action. The flow can then call it with an HTTP POST on the entity endpoint, which fits an existing posting method reached through the OData service.
    • C. Staging class logic belongs to the data management framework and runs as part of a package import. External clients cannot call it, so the flow could not trigger the journal posting on demand.
    • D. Business events are outbound notifications that F&O raises to external subscribers. Subscribing to one cannot invoke a method inside F&O, so the flow still could not start the journal posting.

    6.1: Evaluate data integration patterns and scenarios

    27.An external master data system must be told about every create, update and delete on the vendor entity, whichever form or process changed the record. The team prefers configuration over writing X++ code. What should they do?

    1. A.Create a custom business event for the vendor entity, raised from its insert, update and delete methods, sent to Service Bus
    2. B.Schedule a recurring integration for the vendor entity, exporting its records at short intervals to the master data system
    3. C.Define a virtual entity for vendors, letting the master data system read the table live and compare rows for changes
    4. D.Enable data events for the vendor entity in the business events workspace, publishing to an endpoint such as Event Grid
    Show answer & explanation

    Correct answer: D — Enable data events for the vendor entity in the business events workspace, publishing to an endpoint such as Event Grid

    • A. A custom business event class is X++ code, which the team wants to avoid. It also fires only where a developer raises it, so set-based or bulk changes that bypass the table methods would go unreported.
    • B. A recurring integration runs on a schedule and delivers batches of exported data. The master data system would receive periodic exports rather than a notice of each create, update and delete as it happens.
    • C. A virtual entity gives live read access to the data in Dataverse but publishes no change notifications. The master data system would have to compare rows itself, and deletes would not be reported as events.
    • D. Data events are switched on per entity in the business events workspace and fire on create, update and delete regardless of which form or process made the change. They are published to the chosen endpoint without custom X++.

    6.3: Implement data management

    28.During an import of vendor data, the entity tile in the project shows a warning and no columns are populated in staging. The source file headers are VendorName and VendorGroupCode, but the entity fields are named Name and VendorGroupId. What should the consultant do?

    1. A.Open View map on the entity tile, pair the file columns with the target fields, save the mapping, then rerun the import job
    2. B.Enable Skip staging in the entity setup, point the load at the target tables, save the setup, then rerun the import job
    3. C.Match the casing of the file headers to the staging columns, resave the file, reupload it, then rerun the import job
    4. D.Refresh the entity list in framework parameters, reopen the entity in the project, save it, then rerun the import job
    Show answer & explanation

    Correct answer: A — Open View map on the entity tile, pair the file columns with the target fields, save the mapping, then rerun the import job

    • A. Correct. The file headers (VendorName, VendorGroupCode) differ from the entity field names (Name, VendorGroupId), so no mapping is created on its own. The map has to be edited through View map, saved, and the job rerun.
    • B. Incorrect. Skip staging only changes where rows are written. Source columns must still be mapped to entity fields, and the two sets of names still differ, so the warning on the entity tile stays.
    • C. Incorrect. Changing the case of the headers does not make VendorName match Name or VendorGroupCode match VendorGroupId. The names are different words, and the columns stay unmapped until they are paired in the map.
    • D. Incorrect. Refreshing the entity list reloads the entity definitions. It does not rename the entity fields to follow the file headers or create the column map for this project, so Name and VendorGroupId remain unmatched.

    6.3: Implement data management

    29.Which TWO sources on the job history of a failed import show why a specific entity failed? Select all that apply.(Select 2)

    1. A.The Change tracking option of the entity, which keeps the rejected records and their errors from staging
    2. B.The data package manifest, which holds the validation errors and the rejected rows from the import run
    3. C.The execution details, which show how many records each entity processed successfully and how many failed
    4. D.The Framework parameters page, which lists the error message and the source row of each rejected record
    5. E.The execution log, which includes log text, staging log details with the error column, and the Infolog
    Show answer & explanation

    Correct answers: C, E — The execution details, which show how many records each entity processed successfully and how many failed; The execution log, which includes log text, staging log details with the error column, and the Infolog

    • A. Incorrect. Change tracking on an entity only controls incremental export by detecting changed records. It keeps no list of rejected rows or staging errors from an import job.
    • B. Incorrect. The package manifest describes the entities, their sequence and the mapping in the package. It is not updated with validation errors or rejected rows when a job runs.
    • C. Correct. The execution details show the status of each entity in the job with its successful and failed record counts, and they let you open the staged records to find which entity failed.
    • D. Incorrect. The Framework parameters page holds configuration such as entity execution and staging settings. It does not list error messages or source rows from past runs.
    • E. Correct. The execution log has log text, staging log details that name the error column, and the Infolog for deeper troubleshooting of a failed entity.

    6.4: Integrate with Microsoft Power Platform, Microsoft 365, and Azure

    30.Which statements about finance and operations virtual entities in Dataverse are accurate? (Select all that apply.)(Select 3)

    1. A.Rows remain in the finance and operations database and are read at query time, so Dataverse tables hold no copy.
    2. B.Each call executes the real entity in the app, so validation and business logic on it and its tables are triggered.
    3. C.Placing both environments in the same Azure region keeps the added time for each call below roughly 30 ms.
    4. D.Rows are replicated into Dataverse tables through a dual-write mapping, so queries run against the local copy.
    5. E.Plug-ins running as the Dataverse SYSTEM user can read the finance and operations entities, so no user context is needed.
    6. F.Calls need an ExpressRoute or VPN link between the two environments, so the traffic is encrypted in transit.
    Show answer & explanation

    Correct answers: A, B, C — Rows remain in the finance and operations database and are read at query time, so Dataverse tables hold no copy.; Each call executes the real entity in the app, so validation and business logic on it and its tables are triggered.; Placing both environments in the same Azure region keeps the added time for each call below roughly 30 ms.

    • A. Correct. A virtual entity is a virtual data source: the rows stay in the finance and operations database and are fetched when queried, so Dataverse stores no copy.
    • B. Correct. The adapter service turns Dataverse queries into calls on the actual entity in the app, so the validation and business logic on the entity and its backing tables execute.
    • C. Correct. Microsoft recommends hosting both environments in the same Azure region, which keeps the expected overhead of a virtual entity call under about 30 ms.
    • D. Incorrect. Dual-write is a separate feature that does copy data into Dataverse tables. Virtual entities store no rows in Dataverse, so there is no local copy to query.
    • E. Incorrect. Virtual entity calls need a user context. The SYSTEM user and the out-of-box server-to-server apps do not have access to all entities, and a missing user context causes errors such as Unable to determine the user in CDS.
    • F. Incorrect. Calls already go over a TLS 1.2 secure web call to the CDSVirtualEntityService endpoint, so no ExpressRoute or VPN link is needed for encryption. Cloud environments are supported.

    7: Implement security and optimize performance

    7.2: Apply fundamental performance optimization techniques

    31.A table has an overridden `insert` method that sets derived fields, so `insert_recordset` would be downgraded. To load many rows with fewer round trips while still running that logic, the developer should use `____`.

    1. A.RecordInsertList
    2. B.RecordViewCache
    3. C.SysGlobalObjectCache
    Show answer & explanation

    Correct answer: A — RecordInsertList

    • A. RecordInsertList buffers records and writes them to the database in array inserts while still calling the insert method logic for each record, cutting round trips.
    • B. RecordViewCache caches reads of selected rows for a scope and does not write or batch any inserts to the database.
    • C. SysGlobalObjectCache stores objects in AOS memory across sessions and has no role in inserting table records.

    7.1: Implement security policies and requirements

    32.A new `ContosoBuyer` role must contain everything in the standard purchasing role plus one custom duty, and it must keep receiving Microsoft's later changes to that standard role. What is the best way to build it?

    1. A.Create the new role, nest the standard role inside it as a sub-role, and attach the custom duty to the new role directly
    2. B.Duplicate the standard role in the user interface, add the custom duty to the copy, and name the copy ContosoBuyer
    3. C.Extend the standard role in the custom model, add the custom duty to the extension, and rename it ContosoBuyer
    4. D.Create a privilege with the standard role as its entry point, put it in the custom duty, and add that duty to the new role
    Show answer & explanation

    Correct answer: A — Create the new role, nest the standard role inside it as a sub-role, and attach the custom duty to the new role directly

    • A. Correct: a role can contain other roles as sub-roles. The new role therefore picks up the standard role's duties and privileges, including Microsoft's later updates to them, and the custom duty added to the new role stays separate.
    • B. Incorrect: duplicating a role copies its duties as they are at that moment, and the copy is independent afterwards. Microsoft's later changes to the standard role never reach the copy.
    • C. Incorrect: a role extension adds elements to the existing standard role and cannot be turned into a separate role by renaming it. The custom duty would end up on the standard role itself and affect everyone assigned to it.
    • D. Incorrect: entry points of a privilege are objects such as menu items, actions and service operations, not roles. A role cannot be referenced this way, so the standard role's content is never brought into the new role.

    7.2: Apply fundamental performance optimization techniques

    33.A developer uses `insert_recordset` to copy 300,000 rows into a staging table, but a trace shows one INSERT per row. Which conditions on the target table can cause this downgrade to record-by-record processing? (Select all that apply.)(Select 3)

    1. A.The table's `insert` method is overridden and `skipDataMethods(true)` is not called before the statement
    2. B.Database logging is set up in the system administration forms to record inserts on the staging table
    3. C.An alert rule is set up in the application that triggers when a record is created in the staging table
    4. D.The staging table has several non-unique indexes defined on it, including one on `ItemId`
    5. E.The select that supplies the rows uses an `exists join` against a second table to filter them
    6. F.The staging table's `CacheLookup` property is set to Found, and the application reads it by primary key
    Show answer & explanation

    Correct answers: A, B, C — The table's `insert` method is overridden and `skipDataMethods(true)` is not called before the statement; Database logging is set up in the system administration forms to record inserts on the staging table; An alert rule is set up in the application that triggers when a record is created in the staging table

    • A. Overridden insert code has to run for each record, so the kernel falls back to row-by-row inserts unless skipDataMethods(true) is called before the statement.
    • B. Database logging on inserts needs a log entry written per record, which forces the kernel to process the rows individually instead of as one set-based INSERT.
    • C. An alert on record creation must be evaluated for each new record, so the kernel downgrades the set-based insert to record-by-record processing.
    • D. The database maintains non-unique indexes during a set-based INSERT, so having them, even several, does not cause a downgrade.
    • E. An exists join is a supported part of the select in an insert_recordset statement. It filters the source rows and does not trigger record-by-record processing.
    • F. CacheLookup only controls how reads by key are cached on the application side. It has no bearing on whether the insert runs as a set-based operation.

    7.3: Optimize performance

    34.An ISV solution adds the custom field `CustRiskCode` to `CustTable`. A customer form filters on this field and the query scans the whole table. Which change follows extensibility rules and supports the filter with an index?

    1. A.Create a `CustTable` table extension and add an index on `CustRiskCode` to the Indexes node of that extension.
    2. B.Overlay `CustTable` in the application model and place an index on `CustRiskCode` in the table's own Indexes node.
    3. C.Create a view over `CustTable` and set a clustered index on `CustRiskCode` inside the definition of that view.
    4. D.Set the `CacheLookup` property of `CustTable` to `Found` so filters on `CustRiskCode` are served from the record cache.
    Show answer & explanation

    Correct answer: A — Create a `CustTable` table extension and add an index on `CustRiskCode` to the Indexes node of that extension.

    • A. Correct: a table extension can carry its own indexes, so an index covering the added field is delivered as an extension and the Microsoft-owned table definition stays unmodified.
    • B. Incorrect: an overlay edits the standard table's own metadata. Extensibility rules do not permit overlaying application code in new solutions, so the index has to be delivered through an extension.
    • C. Incorrect: a view has no storage of its own, so an index setting on it does nothing for the base table. Queries against CustTable still scan the whole table.
    • D. Incorrect: CacheLookup governs record caching, which serves lookups on a unique key. It creates no index and does not speed up filtered queries on a non-key field, so the scan remains.

    7.3: Optimize performance

    35.A nightly import of several million rows through Data management runs in one batch task and takes hours. The team wants it split into parallel tasks across the batch servers. Where is this configured?

    1. A.In Framework parameters, set `Import threshold record count` and `Import task count` per entity to size the work units.
    2. B.In Server configuration, raise `Maximum batch threads` on the batch servers to let the import task use more threads.
    3. C.In Batch groups, add the batch servers to the group assigned to the import job to let it spread across them.
    4. D.In Data entities, turn on `Set-based processing` for the entity to run the import as bulk database operations.
    Show answer & explanation

    Correct answer: A — In Framework parameters, set `Import threshold record count` and `Import task count` per entity to size the work units.

    • A. Correct: the entity execution parameters in Framework parameters set how many records go into each import task and how many tasks are created, so a large import becomes several tasks that batch servers pick up.
    • B. Incorrect: Maximum batch threads only limits how many tasks a batch server runs at the same time. The import is still one task, so a higher thread count does not divide it.
    • C. Incorrect: batch groups decide which servers may run a job's tasks. The import is still a single task, so adding servers to its group gives them nothing to share.
    • D. Incorrect: set-based processing changes how rows are written, using set operations in the database. It does not create multiple import tasks, so the work is not spread over the batch servers.

    Want the full experience?

    These are just samples. Practice the full Microsoft Dynamics 365: Finance and Operations Apps Developer (MB-500) - December 18 onwards question bank in quiz mode — free, no signup, with domain practice and exam simulation.