You are tasked with loading data from a set of CSV files located in an Amazon S3 bucket into a Snowflake table named 'CUSTOMER DATA'. The files have headers, but some files might have inconsistent col

Question

Options

• AUse along with 'ON_ERROR = CONTINUE in the COPY INTO statement. Define a file format
• BUse a simple 'COPY INTO' statement without any specific options and rely on Snowflake's
• CCreate a view on top of the staged data using 'SELECT statements that explicitly map columns by
• DUse 'ORDER BY clause in the 'COPY INTO' statement to explicitly define the order of columns
• EDefine a file format object with `SKIP_HEADER = 1' and -- (if applicable). Use 'MATCH BY

• A
  75% (21)
• B
  7% (2)
• C
  4% (1)
• D
  14% (4)

Explanation

Option A is correct because it combines the two critical features needed for this scenario: MATCH_BY_COLUMN_NAME (implied by the option's context) handles the inconsistent column ordering by using CSV header names to map columns to the table schema automatically, while ON_ERROR = CONTINUE ensures the load job proceeds even if individual rows or files have issues - together minimizing manual intervention and maximizing data throughput.

Why the distractors are wrong:

• B - A plain COPY INTO without options assumes positional column matching; inconsistent column order will silently load data into wrong columns or error out.
• C - Creating a SELECT-based view with explicit column mapping works but requires manual mapping for every schema variation, defeating the goal of minimizing intervention.
• D - ORDER BY is not a valid clause in COPY INTO; it belongs to query syntax, not data loading commands.
• E - While SKIP_HEADER = 1 and MATCH_BY_COLUMN_NAME are individually correct tools, this option omits the ON_ERROR = CONTINUE safety net needed to ensure all relevant data loads despite inconsistent files.

Memory tip: Think "Name it, Don't Order it, Don't Stop it" - use MATCH_BY_COLUMN_NAME so Snowflake matches by name not position, and ON_ERROR = CONTINUE so it doesn't stop on bad rows. Any option that relies on column position or manual mapping is wrong for variable-order CSVs.

No community discussion yet for this question.