MB-700 flashcards

32 cards. Read the term and the definition here. The deck at the top of this page is optional.

Environment reset capability

Only Sandbox environments can be reset to a point-in-time backup; Production environments cannot be reset, so destructive testing belongs in Sandbox copies.

Managed vs unmanaged solutions

Unmanaged solutions allow direct editing and are used in development; managed solutions are locked, upgradable, and are the only supported artifact for downstream environments.

Solution patch constraint

A patch can only be created from a managed parent solution in the same environment, and it clones the parent while holding only the delta changes.

Connection reference purpose

A connection reference is a solution-aware pointer to a connector connection, letting each environment bind its own credentials instead of embedding them in the flow definition.

Environment variable purpose

Environment variables store environment-specific parameters such as URLs, IDs, or feature flags inside a solution, with optional per-environment current values and defaults.

Virtual table use case

Virtual tables surface external data in Dataverse without replication, are read-mostly, and do not support the full feature set of native tables such as offline or complex relationships.

Business unit data partitioning

Business units partition record ownership; a user in one business unit cannot see another business unit's records unless a security role grants a higher access level or a team shares them.

Security role access levels

Privileges combine a permission (Create, Read, Write, Delete, Append, Assign, Share) with an access level: User, Business Unit, Parent:Child Business Unit, or Organization.

App type selection rule

Choose model-driven for Dataverse-centric role-based processes, canvas for connector-heavy task apps and mobile use, and Power Pages for external or anonymous audiences.

Publisher prefix decision

The publisher prefix is baked into every schema name at creation and cannot be changed afterward, so settle the prefix before any table or column is created.

Fit-gap classification

Each requirement is classified as standard configuration, extension through code or custom components, or a process change; extensions carry the highest upgrade and testing cost.

Choices vs lookup tables

Use choices for small static option sets that rarely change; use a lookup table when values are administered by users, need metadata, or exceed a manageable list size.

Business process flow limitation

A business process flow guides stage progression in the UI but does not enforce data integrity; API writes and other clients can bypass it, so validation needs plugins or rules.

Sync vs async plugin rule

Synchronous plugins run inside the save transaction and should stay fast and local; asynchronous plugins run in the background and suit integrations, long calculations, and retries.

Data migration staging pattern

Legacy data is loaded into staging tables, transformed, then imported; disable or filter plugins enforcing business rules during the load to avoid failures and slowdowns.

Licensing model decision

Per-user licenses cover unlimited apps for a named user; per-app licenses cover one app for many users; Power Pages and Dataverse capacity are licensed separately from user access.

Reporting architecture choice

Use Dataverse views and dashboards for in-app operational reporting, and Power BI with import or DirectQuery for analytical reporting across large or joined datasets.

Non-functional requirements to capture

Record expected data volumes, peak concurrency, data residency and compliance constraints, recovery time and recovery point objectives, and offline requirements before designing the solution.

Custom table cost rule

Every custom table adds schema maintenance, security role entries, and reporting overhead; prefer standard tables when semantics match and extend them with custom columns.

Offline capability decision

Model-driven apps support mobile offline profiles for field scenarios; canvas apps rely on their own offline functions and local collections, which changes the data conflict strategy.

Solution packager and source control

The solution packager extracts a solution zip into XML files that can be diffed and merged in Git, enabling branch policies and reviewable schema changes.

Deployment stage promotion order

Development holds the unmanaged solution; build, test, UAT, and production receive managed imports, with an approval gate before production and a backup taken first.

Unresolved connection reference symptom

If a connection reference has no bound connection in the target environment, flows fail to activate on import; bind connections per environment before the deployment window.

DLP policy risk

Tenant data loss prevention policies can block a connector used by a solution, breaking flows at runtime; confirm the business/data connector grouping before go-live.

API request entitlement consumption

Dataverse API requests are consumed by app usage, integrations, and load tests; exhausting the entitlement throttles the environment, so load tests need a dedicated environment.

Work breakdown by component

Decomposing work by solution component such as tables, apps, flows, and integrations enables parallel workstreams without two teams editing the same solution simultaneously.

Change control board role

The change control board approves scope changes, and each approved change is logged as a backlog item traceable to a requirement and a test case.

Rollback plan requirement

A deployment runbook must state how to restore the pre-deployment backup and which solution version to reimport if a managed upgrade fails in production.

Test pyramid for Power Platform

Unit tests cover plugin and PCF code with mocked contexts, integration tests cover flows and APIs, and end-to-end UI tests are fewest because they are slowest.

UAT entry criteria

UAT starts only when the managed solution is deployed to a refreshed UAT environment, critical test cases pass in test, and business has signed off the scripts.

Shared component regression rule

Changing a shared component such as a table column, security role, or connection reference requires regression testing of every app, flow, and report that consumes it.

Go-live exit criteria

Go-live requires zero open blocker defects, an approved deferral list for non-blockers, a completed cutover rehearsal, and a signed rollback plan.