What is the most likely cause if a large join query takes hours to complete even after increasing the warehouse size?

• A. The query was still running before the new size of the warehouse was applied
• B. The warehouse had auto-suspend turned on and restarted mid-query
• C. The warehouse doesn't have enough power still
• D. One of the column's values is significantly more than the rest of the values in the column; that's why it produces a skew in your data.

Answer: (D)

When a join processes a lot of data, uneven work distribution across compute resources can cause hours-long runtimes. If one value in a join key is Extremely common, that value ends up matching a large portion of the data, so one or a few nodes have to do most of the heavy lifting while the rest sit idle. Even larger warehouses increase parallelism, but the overall time is still dominated by that slowest worker handling the skewed key. That’s why data skew in the join key is the most plausible reason for the persistent long runtime.

To verify, check the distribution of the join key values in both tables and identify any values that occur far more often than the rest. If skew is present, you can address it by filtering out the outliers early, restructuring the query to reduce the amount of data joined (perhaps by aggregating first or performing the join in stages), or adjusting clustering/partitioning strategies to improve how data is pruned before the join.