DP-900 Describe an analytics workload on Azure Practice Question

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

• ✓

  Create the Stream Analytics job, then configure inputs, then configure outputs, then define the transformation query.

  Why this is correct

  You must first create the Stream Analytics job as an Azure resource, because the job is the container that owns all other components. After the job exists, you configure inputs (such as Event Hubs or IoT Hub) and outputs (such as Blob storage or Power BI), because these are attached to the job and expose the schemas and sinks your query will use. Finally, you define the transformation query in the job's Query editor, where you reference the input and output aliases. This sequence ensures the query can be validated against real input and output metadata before you start the job.

• ✗

  Create the Stream Analytics job, then define the transformation query, then configure inputs, then configure outputs.

  Why it's wrong here

  Creating the job first is correct, but defining the transformation query before configuring inputs and outputs is not possible in a functional way. The query references input and output aliases in its FROM and INTO clauses, and those aliases must already exist as configured sources and sinks so the service can validate field names, types, and query syntax. Without inputs and outputs, the query editor shows unresolved identifiers, and the job cannot be started. Therefore, this ordering fails because the query depends on metadata that has not yet been defined.

• ✗

  Configure inputs, then configure outputs, then create the Stream Analytics job, then define the transformation query.

  Why it's wrong here

  Inputs and outputs cannot be configured first because they are child resources of the Stream Analytics job and require the job's resource ID and scope to be created. The Azure portal, REST API, and ARM templates all require an existing job before you can add an input or output; attempting to do otherwise leaves no parent resource to map the data source to. Even if you could somehow define them, a query defined later would reference aliases without a validated job context. This reverses the dependency hierarchy, so the job must be the first resource created.

• ✗

  Define the transformation query, then create the Stream Analytics job, then configure inputs, then configure outputs.

  Why it's wrong here

  Defining the transformation query before creating the job is impossible because there is no job resource to host the query editor, and the query cannot exist as a standalone artifact. Even if you were to create the job afterward, the query would still be invalid because it requires input and output aliases to be configured, and at the time the query is written those aliases do not exist and cannot be resolved to any actual data source or sink. This option places every dependent step before its prerequisites, making it the least viable sequence. The only valid starting point is creating the Stream Analytics job itself.