Understand the Features of Liquid Clustering and Predictive Optimization. DATA-ENGINEER-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

1. 6application · medium

   A company has an existing Delta table that is currently partitioned by `year`. They want to improve query performance for queries that filter on `customer_id` and `region`. They do not want to change the existing partition scheme. What is the correct way to enable Liquid Clustering on this table?

   • ARun `OPTIMIZE table_name CLUSTER BY (customer_id, region)`.
   • BRecreate the table with `PARTITIONED BY (customer_id, region)`.
   • CRun `CREATE TABLE table_name CLUSTER BY (customer_id, region)`.
   • DRun `ALTER TABLE table_name CLUSTER BY (customer_id, region)`.
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A data engineer is designing a Delta table for a social media analytics platform. The table will store user engagement events and will be queried by `user_id`, `event_type`, and `event_date`. The team wants to optimize for queries that filter on `user_id` and `event_date`. What should the engineer consider when choosing clustering keys?

   • AUse `user_id` and `event_date` as clustering keys because they are commonly used in query filters.
   • BUse all three columns as clustering keys to cover all possible queries.
   • CDo not use clustering keys because the table is small.
   • DUse `event_type` as the only clustering key because it has low cardinality.
   Reveal Solution Discussion Select an answer first
4. 8application · medium

   A company has a Delta table that is updated with new data every hour. They currently use Z-ordering on a column to improve query performance, but they find that they need to run OPTIMIZE frequently to maintain the layout. They want a solution that requires less manual maintenance. What should they do?

   • ASwitch to Liquid Clustering with the same column as the clustering key.
   • BContinue using Z-ordering but run OPTIMIZE more often.
   • CDisable all clustering and rely on file compaction only.
   • DPartition the table by the column used for Z-ordering.
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   How do clustering keys enable data skipping in Delta Lake?

   • ABy physically sorting all data in the table by the clustering keys, so the engine can binary-search for matches
   • BBy partitioning the table into separate directories for each clustering key value
   • CBy creating a separate index for each clustering key column that is updated on every write
   • DBy storing statistics about the clustering key values in the transaction log, allowing the query engine to skip files that do not match the query filter
   Reveal Solution Discussion Select an answer first
6. 10foundation · easy

   How does Liquid Clustering differ from traditional partitioning?

   • ALiquid Clustering automatically partitions data by date, while partitioning requires manual specification
   • BLiquid Clustering is only available for tables created with Databricks Runtime 15.0 or later
   • CLiquid Clustering allows clustering keys to be changed without rewriting the entire table, while partitioning requires a full rewrite to change partition columns
   • DLiquid Clustering requires a fixed number of clustering keys, while partitioning allows any number of partition columns
   Reveal Solution Discussion Select an answer first