Managing data distribution
Distribution style is a table property that dictates how that table’s data is distributed throughout the compute nodes. The goal of data distribution is to leverage the massively parallel processing of Amazon Redshift and reduce the I/O during query processing to improve performance. Amazon Redshift Advisor provides actionable recommendations on distribution style for the table with the alter statement. Using automatic table optimization enables you to self-manage the table distribution style based on workload patterns:
- KEY: The value is hashed; the same value goes to the same location (slice)
- ALL: All table data goes to the first slice of every compute node
- EVEN: Round-robin data distribution across the compute nodes and slices
- AUTO: Applies EVEN, ALL, and KEY distribution based on the scenario
Figure 8.5 – Data distribution styles
In this recipe, you will see how Amazon Redshift’s automatic...
