A financial services company is using Snowflake Streams on a table 'TRANSACTIONS' to capture changes for auditing purposes. The 'TRANSACTIONS' table contains sensitive data, and the auditing team requires the stream to only capture changes to specific columns: 'ACCOUNT ID', 'TRANSACTION DATE', and 'TRANSACTION AMOUNT'. Which of the following approaches is the MOST efficient and secure way to achieve this requirement, ensuring minimal performance impact and data exposure?

• A. Create a standard Stream on the 'TRANSACTIONS table and then filter the results in downstream processing to only include the required columns.
• B. Create a Stream on the 'TRANSACTIONS' table and use a masking policy on the stream's output to redact the unnecessary columns.
• C. Create a View that selects only the 'ACCOUNT ID, 'TRANSACTION DATE, and 'TRANSACTION AMOUNT columns and create a Stream on the View.
• D. Create a task that clones the TRANSACTIONS table and a stream on that cloned table, limiting what changes are captured using a WHERE clause on the cloning command.
• E. Create a Stream on the 'TRANSACTIONS' table. Periodically truncate stream and reload all data from TRANSACTION table by applying filter while loading.

A Snowflake data pipeline ingests data from multiple external sources into a RAW DATA table. A transformation process then moves the data to a ANALYTICS DATA table, applying several complex UDFs written in Java and Python for data cleansing and enrichment. Performance is significantly slower than expected. Which combination of techniques would BEST improve the performance of this transformation pipeline?

• A. Rewrite the UDFs in SQL or Snowpark Python/Java for better integration with the Snowflake engine and leverage vectorization where possible; cache intermediate results using temporary tables.
• B. Reduce the number of UDF calls by consolidating them into a single, more complex UDF. Replace the transformation pipeline with a series of COPY INTO statements.
• C. Increase the virtual warehouse size and re-cluster the ANALYTICS DATA table based on the most frequently filtered columns.
• D. Implement data partitioning in the RAW DATA table based on ingestion time and switch to using stored procedures instead of transformation pipelines.
• E. Use external functions instead of UDFs to offload the processing to an external compute environment and configure auto-scaling for the virtual warehouse.

Consider a scenario where you have a Snowflake external table 'ext_logs' pointing to log files in an S3 bucket. The log files are continuously being updated, and new files are added frequently. You want to ensure that your external table always reflects the latest data available in S3. Which of the following actions and configurations are required or recommended to keep the external table synchronized with the underlying data source? (Select all that apply)

• A. Enable auto-refresh on the external table using the 'AUTO_REFRESH = TRUE parameter during creation.
• B. Periodically execute the 'ALTER EXTERNAL TABLE ext_logs REFRESH' command to update the metadata about the files in S3.
• C. Create a Snowpipe that continuously loads data from the S3 bucket into a Snowflake table instead of using an external table.
• D. Implement a stored procedure that periodically executes a query on the external table to force a metadata refresh.
• E. Configure an event notification service (e.g., AWS SQS) to trigger an external table refresh whenever new files are added to S3.

You need to implement both a row access policy and a dynamic data masking policy on the 'EMPLOYEE table in Snowflake. The requirements are as follows: 1. Employees should only be able to see their own record in the 'EMPLOYEE table. 2. The 'SALARY' column should be masked for all employees except those with the 'HR ADMIN' role. Unmasked values are required for compliance reasons, they need to be available for 'HR ADMIN' role. Given the following table structure: CREATE TABLE EMPLOYEE ( EMPLOYEE ID INT, EMPLOYEE NAME STRING, SALARY NUMBER, EMAIL STRING ) ; Which of the following sets of steps correctly implement the row access policy and dynamic data masking policy?

• A. Option A
• B. Option C
• C. Option E
• D. Option B
• E. Option D

A data engineer notices that a daily ETL job loading data into a Snowflake table 'TRANSACTIONS' is consistently taking longer than expected. The table is append-only and partitioned by 'TRANSACTION DATE. The engineer observes high 'Remote Spill' during the load process and suspect that micro-partition pruning isn't working effectively. Which of the following approaches would BEST address the performance issue, assuming you have already considered increasing warehouse size?

• A. Partition the data in the source system by 'TRANSACTION DATE' and load data in parallel corresponding to each partition.
• B. Examine the data load process to ensure the data is loaded in 'TRANSACTION_DATE order. If not, sort the data by 'TRANSACTION_DATE before loading.
• C. Enable automatic clustering on the 'TRANSACTION_DATE column of the 'TRANSACTIONS table.
• D. Re-create the 'TRANSACTIONS' table with a larger virtual warehouse and re-load the entire dataset.
• E. Implement data skipping by creating a masking policy on the 'TRANSACTION_DATE column.