Question 37
Domain 10: Data ModellingA data engineering team wants to improve query performance on a large table, but the table's access patterns change over time and they want to avoid the rigidity of predefined partition columns or repeated ZOrder maintenance. Which approach best fits this requirement?
Correct answer: A
Explanation
Liquid Clustering is preferable when clustering needs may evolve, because it provides more flexibility than fixed Partitioning and can reduce the maintenance burden associated with ZOrder. — official.txt
Why each option is right or wrong
A. Use liquid Clustering because it is more adaptable than Partitioning and avoids relying on repeated ZOrder optimization.
The source material frames liquid Clustering as having comparative advantages over both Partitioning and ZOrder. In this scenario, the team specifically needs flexibility as query patterns change and wants to avoid the rigidity of predefined partition columns and ongoing ZOrder-style maintenance, which matches liquid Clustering.
B. Use Partitioning because fixed partition columns are the best choice when query patterns frequently change.
Partitioning uses predefined partition columns, so it is less suitable when access patterns frequently change.
C. Use ZOrder because it replaces the need to consider evolving clustering strategies over time.
ZOrder is contrasted with liquid Clustering, which is presented as the more flexible approach for changing needs.
D. Keep the table unclustered because neither Partitioning nor ZOrder has tradeoffs compared with liquid Clustering.
The material explicitly discusses comparative advantages of liquid Clustering over Partitioning and ZOrder.