SOL-C01 Question #205: Real Exam Question with Answer & Explanation

Question

You are tasked with classifying a large dataset of customer support tickets using Snowflake Cortex's CLASSIFY TEXT function. You need to ensure optimal performance and cost- effectiveness. Which of the following strategies would be MOST effective?

Options

• AUse a very large warehouse (e.g., X-LARGE) to execute the CLASSIFY _ TEXT function,
• BIncrease the parameter to a very high value to prevent query failures due to the long processing
• CProfile the execution time of CLASSIFY _ TEXT with different warehouse sizes and choose the
• DDisable auto-suspension on your warehouse to keep it running continuously while classifying the
• ELoad all customer support tickets into a single very wide row and classify that single row using

Explanation

Option C is the most sensible approach. Profiling with different warehouse sizes allows you to determine the optimal balance between performance and cost. Using a materialized view ensures that the classification is only performed once, and subsequent queries are faster. Option A is wasteful, as a very large warehouse may not be necessary. Option B addresses timeout issues but doesn't optimize performance or cost. Option D is also wasteful and can lead to unnecessary costs. Option E is illogical as CLASSIFY _ TEXT is applied to text and not a representation of the entire dataset in a single row.

No community discussion yet for this question.