Free PL-500 practice test: 30 questions with answers

A child flow is a reusable component that receives inputs from a parent and returns outputs, which helps structure larger automations. Copying, deleting, and notifying are not the purpose of a child flow.

Concurrency control limits how many runs or loop iterations execute in parallel, which protects downstream systems. Display language, connector counts, and package size do not control parallelism.

A retry policy re-attempts a failed action a set number of times with an interval, which handles transient errors. Moving runs to the next morning, removing owners, and printed logs are not retry policies.

Cloud flows use workflow expression functions such as concat, parseJson, and formatDateTime to transform values. Spreadsheet formulas, processor instructions, and printer codes are not the expression language of cloud flows.

Datatables are in-memory tables that the flow can read, write, and loop over row by row. Printed results, restricted database tables, and wallpaper graphics do not describe datatables.

A subflow groups reusable actions that the main flow calls, which keeps the design modular. Duplicate copies, hidden logic, and output storage do not describe subflows.

The exception handling block catches system exceptions so the flow can respond cleanly. Variables, loops, and printing blocks do not handle system errors.

Secure variables mask and protect sensitive values such as passwords in logs and run history. Network-wide visibility, public text files, and scheduled changes weaken protection rather than provide it.

A trigger condition is an expression that must be true before a run starts, which filters out events that should not start the flow. Button colours, time zones, and history retention are different settings.

Both cloud flows and desktop flows can parse JSON, with cloud flows using a schema-driven Parse JSON action. Claims that JSON cannot be parsed, that XML must be read by hand, and that printed copies are needed are incorrect.

Desktop flows run on Windows machines and automate application user interfaces, which is the RPA capability the question describes. Cloud flows, business process flows, and solution packages serve different purposes.

An attended run happens while a user is present on the machine, which suits tasks that need the person's involvement. Unattended runs, remote virtual machines, and connector-only actions describe different execution modes.

An automated trigger responds to an event in a connected service, such as a new email. Manual, scheduled, and instant triggers start flows in different ways that do not depend on the mailbox event itself.

Unattended runs suit processes that should run without a person present, such as overnight batches on a registered machine. Step-by-step human decisions, phone-only browsers, and per-step approvals suit attended or cloud-based designs instead.

The HTTP action sends a request to an external REST endpoint and receives the response, which is how a cloud flow exchanges data with a web service. Printing, browser colours, and installing browsers are unrelated.

Work queues suit high volumes of independent items that need retries and status tracking. Dashboard colours, break counts, and font choices do not affect whether a queue is the right design.

Custom connectors are justified when a needed API is not in the catalogue or when a private API must be called consistently. Designer colours, bypassing the API, and messaging without a licence do not justify building one.

Power Apps builds canvas and model-driven apps that use connectors and Dataverse as data sources. Power BI, Azure DevOps, and Forms serve visualization, delivery, and surveys rather than building apps on Dataverse.

A business process flow guides users through stages of a process inside a model-driven app, with steps and fields for each stage. Background file moves, desktop clicking, and printed checklists are different things.

The on-premises data gateway lets cloud flows reach data inside the network through a SQL Server connector with proper authentication. Opening ports to the internet, USB copies, and manual retyping are unsafe or unworkable approaches.

Machine groups allow desktop flows to be load balanced across registered machines, often together with queues. Email groups, physical room grouping, and printer lists are not machine group functions.

Machine registration connects a Windows machine to the service so it can run and be managed for desktop flows. Asset register entries, user accounts, and operating system installs are separate activities.

DLP policies classify connectors and restrict combinations, such as business and non-business connectors in one flow. Preventing all saves, deleting history, and print encryption are not DLP functions.

Solutions package related components so they can be transported between environments in an ALM process. Vendor selection, password storage, and printed design documents are not solution functions.

A dedicated service account keeps unattended automation independent of any person and makes its access manageable and auditable. Using personal mailboxes, shutdowns, and priority claims are not reasons for service accounts.

The run queue shows the status of queued items, such as waiting, running, or completed, for desktop flow processing. Visitor lists, cup counts, and printer data are unrelated.

Sharing a flow grants the chosen role, such as viewing or running the flow, to another user. It does not transfer machine ownership, bank access, or the source of other flows.

Connectors in the Blocked group cannot be used in flows in the covered environments, which stops data leaving through them. Administrator-only access, desktop-only use, and hidden-but-usable connectors describe other configurations.

Connections hold the credentials that flows use, so after a password change the connection must be updated to keep runs authenticated. Leaving the old password, printer drivers, and wallpapers do not affect authentication.

An environment is a container for apps, flows, and data that keeps development, test, and production separate. Office rooms, print virtual machines, and switch lists are unrelated.