DP-200 Test 1 — Questions and Answers
Question 1: You have an Azure Data Factory pipeline that loads source data into an Azure Data Lake Storage Gen2 account incrementally. <br> <br> The source table's LastUpdatedDate column identifies the data to be loaded. <br> <br> Every four hours, you intend to run the pipeline. <br> <br> You must make certain that the pipeline's execution fits the following criteria: <br> <br> When a pipeline run fails owing to concurrency or throttling constraints, the execution is automatically retried. <br> <br> Backfilling existing data in the table is supported. <br> <br> What kind of trigger should you employ?
- schedule
- event
- tumbling window (Correct answer)
- on-demand
Correct answer: tumbling window
A tumbling window trigger is ideal for this scenario because it fires at a fixed time interval (every four hours) and creates non-overlapping, contiguous windows. This trigger type inherently supports automatic retries for failed executions and allows for easy backfilling of existing data by re-running specific past windows. These features ensure robust and manageable incremental data loading.
Question 2: Pool1 is an Azure Synapse Analytics dedicated SQL pool, and DB1 is a database. Table1 is a fact table found in DB1. <br> <br> In Table1, you must determine the amount of the data skew. <br> <br> In Synapse Studio, what should you do?
- Connect to Pool1 and run DBCC CHECKALLOC
- Connect to the built-in pool and query sysdm_pdw_sys_info
- Connect to Pool! and query sys.dm_pdw_nodes_db_partition_stats (Correct answer)
- Connect to the built-in pool and run DBCC CHECKALLOC
Correct answer: Connect to Pool! and query sys.dm_pdw_nodes_db_partition_stats
To determine data skew in an Azure Synapse Analytics dedicated SQL pool, you should connect to the dedicated SQL pool (Pool1) and query the `sys.dm_pdw_nodes_db_partition_stats` Dynamic Management View (DMV). This DMV provides detailed information about data distribution across the compute nodes, including row counts per partition, which is crucial for identifying uneven data distribution or skew. Data skew can significantly impact query performance in a distributed system like Synapse.
Question 3: Azure Stream Analytics is used to receive Twitter data from Azure Event Hubs and send it to an Azure Blob storage account. <br> <br> Every minute, you must output the number of tweets from the previous five minutes. <br> <br> Which windowing mode should you employ?
- Session
- Hopping (Correct answer)
- Sliding
- Tumbling
Correct answer: Hopping
A hopping window is the correct choice because it allows you to define a fixed-size window (five minutes) that moves forward by a specified 'hop' interval (every minute). This enables the system to calculate the number of tweets in the preceding five minutes, and then perform the same calculation one minute later, effectively providing a rolling five-minute count every minute. Tumbling windows are non-overlapping, and sliding windows output for every event, neither of which fits the requirement precisely.
Question 4: You're working on an Azure Data Lake Storage Gen 2 application. <br> <br> You must propose a method for granting permissions to a certain application for a set length of time. <br> <br> What should your advice include?
- shared access signatures (SAS) (Correct answer)
- Azure Active Directory (Azure AD) identities
- role assignments
- account keys
Correct answer: shared access signatures (SAS)
Shared Access Signatures (SAS) are the recommended method for granting time-limited and granular permissions to specific resources in Azure Data Lake Storage Gen2. A SAS token provides delegated access to storage resources for a specified period, with defined permissions and an optional IP address restriction. This ensures that the application only has the necessary access for a controlled duration, enhancing security.
Question 5: You're utilizing Azure metrics to keep track of an Azure Stream Analytics task. <br> <br> You notice that the average watermark delay over the last 12 hours has consistently exceeded the configured late arrival tolerance. <br> <br> What could be the source of this behavior?
- The input data contains errors.
- Events with an application timestamp that is more than five minutes earlier than their arrival time are received as inputs.
- The job's resources are insufficient to handle the volume of incoming data. (Correct answer)
- The late arrival policy causes events to be dropped
Correct answer: The job's resources are insufficient to handle the volume of incoming data.
A consistently high average watermark delay exceeding the late arrival tolerance indicates that the Azure Stream Analytics job is struggling to process incoming data at the rate it's arriving. This usually points to insufficient resources allocated to the job. Increasing the job's streaming units (SUs) would provide more processing power, allowing it to catch up with the incoming event stream and reduce the watermark delay.
Question 6: You have a virtual network that protects your Azure Data Lake Storage Gen2 account, adls2. <br> <br> In Azure Synapse, you're creating a SQL pool with adls2 as a source. <br> <br> What method should you use to connect to adls2?
- a shared key
- a shared access signature (SAS)
- an Azure Active Directory (Azure AD) user
- a managed identity (Correct answer)
Correct answer: a managed identity
When connecting an Azure Synapse SQL pool to an Azure Data Lake Storage Gen2 account that is protected by a virtual network, using a managed identity is the most secure and recommended approach. A managed identity provides an Azure Active Directory identity for the Synapse workspace, which can then be granted specific permissions to the ADLS Gen2 account. This eliminates the need to manage credentials manually and allows for secure, service-to-service authentication within the Azure ecosystem, especially with VNet integration.
Question 7: In Azure Synapse Analytics, you have a data warehouse. <br> You must ensure that c rests. <br> What do you want to enable?
- Advanced Data Security for this database
- Transparent Data Encryption (TDE) (Correct answer)
- Secure transfer required
- Always Encrypted for all columns
Correct answer: Transparent Data Encryption (TDE)
Transparent Data Encryption (TDE) is the standard method for ensuring data at rest is encrypted in Azure Synapse Analytics dedicated SQL pools. TDE encrypts the entire database, including data files and log files, on the storage layer. This provides a crucial layer of security by protecting data from unauthorized access even if the underlying storage media is compromised.
Question 8: You have a subscription to Azure that includes the following resources: <br> * A SQL API account for Azure Cosmos DB <br> * An encryption key named key1 is stored in an Azure key vault named vault1. <br> By utilizing key1, you must verify that all content stored in Azure Cosmos DB is encrypted. What should you start with?
- Create a managed identity (Correct answer)
- Create a new Azure Cosmos DB account
- Disable soft delete on vault 1
- Enable geo-redundancy for the Azure Cosmos DB account
Correct answer: Create a managed identity
To encrypt Azure Cosmos DB data using customer-managed keys (CMK) from Azure Key Vault, the Cosmos DB account first needs an identity that can access the Key Vault. Creating a managed identity (either system-assigned or user-assigned) for the Azure Cosmos DB account is the initial and essential step. This managed identity can then be granted the necessary permissions (Get, Wrap Key, Unwrap Key) on the encryption key in Azure Key Vault, allowing Cosmos DB to use it for data encryption.
Question 9: You're keeping an eye on an Azure Stream Analytics job. <br> For the last hour, the number of Backlogged Input Events has been 20. <br> The number of Backlogged Input Events must be reduced. <br> What are your options?
- Stop the job
- Drop late arriving events from the job
- Add an Azure Storage account to the job
- Increase the streaming units for the job (Correct answer)
Correct answer: Increase the streaming units for the job
A persistent number of backlogged input events in an Azure Stream Analytics job indicates that the job's processing capacity is insufficient to handle the incoming data volume. To reduce this backlog and allow the job to catch up, you should increase the streaming units (SUs) allocated to the job. Streaming units represent the computing resources, and more SUs provide greater throughput and processing power.
DP-200 Test 1
You have an Azure Data Factory pipeline that loads source data into an Azure Data Lake Storage Gen2 account incrementally.
The source table's LastUpdatedDate column identifies the data to be loaded.
Every four hours, you intend to run the pipeline.
You must make certain that the pipeline's execution fits the following criteria:
When a pipeline run fails owing to concurrency or throttling constraints, the execution is automatically retried.
Backfilling existing data in the table is supported.
What kind of trigger should you employ?