Master the Snowflake SnowPro Core Certification: Free Practice Test – 1

The SINGLE=TRUE option ensures that all the data is written into a single file instead of multiple files when unloading data using COPY INTO. This guarantees that the output file will have the exact name specified by the user.

The file_extension option must be used to specify the exact file extension (e.g., .csv) for the unloaded file. By default, if the SINGLE option in the COPY INTO command is set to TRUE, Snowflake unloads a single file without a file extension.

Snowflake recommends dedicating a separate warehouse for loading data when planning virtual warehouse usage for data loads. This helps optimize performance by isolating the data load processes from query workloads, ensuring that neither is impacted by the other. Using a dedicated warehouse also allows better control over resource allocation and helps maintain efficient data processing without interference.

When performing a FLATTEN command on semi-structured data in Snowflake, the resulting columns typically include KEY, which represents the name of the attribute, and VALUE, which represents the corresponding data. These are essential for working with hierarchical or nested data structures like JSON.

This parameter controls the maximum length of time a query can remain in a queued state when there is contention for warehouse resources before being canceled. It helps manage the load during high-demand periods by setting limits on how long a query will wait to be processed.

Snowflake stores a table's underlying data in a columnar file format, which optimizes performance for analytical queries, and organizes the data into micro-partitions, which are automatically managed by Snowflake to improve query performance and reduce storage costs.

A Snowflake stream object is used to track DML changes (inserts, updates, and deletes) made to a table. It allows users to query the change data and process it as needed, facilitating use cases such as change data capture (CDC).

The same exact query will return the precomputed results if the underlying data hasn't changed and the results were last accessed within the previous 24-hour period, and the 24-hour timer on the precomputed results gets renewed every time the exact query is executed.

Parquet is the file format that will keep floating-point numbers from being truncated when data is unloaded. Parquet is a columnar storage format that efficiently handles complex data types, including floating-point numbers, preserving their precision during data storage and retrieval.

Snowflake tables are logical representations of underlying physical data, meaning they abstract the physical storage and are managed by Snowflake automatically.