SOL-C01 Question #294: Real Exam Question with Answer & Explanation

Question

A Snowflake Notebook is configured to run a series of data processing steps on a schedule using the Snowflake Task feature. One of the steps involves calling an external API to enrich the data. Due to network instability, the API calls occasionally fail, causing the entire task to fail. You want to implement error handling within the notebook to retry failed API calls up to a certain number of times before giving up. Assuming you are using Python within the Snowflake Notebook and Snowpark, how would you best implement this retry mechanism while ensuring minimal disruption to the notebook's workflow and preserving error information?

Options

• AWrap the API call within a 'try...except' block. Inside the `except' block, use `time.sleep()' to
• BImplement a custom decorator function that handles retries. The decorator should catch
• CUse the `snowflake.snowpark.functions.call_udf function with the 'retry_count' parameter set to the
• Dwithin notebooks. A decorator function (B) can work, but `tenacity' provides more features with
• EUse a dedicated Python library like 'tenacity' within the Snowflake Notebook. Configure `tenacity' to

Explanation

Using a dedicated library like tenacity' (E) provides the most robust and flexible retry mechanism. It offers features like exponential backoff, customizable retry strategies, and detailed logging. While try...except' blocks (A) can work, they are less elegant and harder to maintain for complex retry logic. Snowflake does not have direct error handling mechanisms for external API calls (C,

No community discussion yet for this question.