When we discuss data warehouse in class using AWS Redshift as the example, it is pointed out that a data warehouse system is running in a cluster of computers with a lead computer node working as the interface between clients $($such as SQL clients$)$ and compute nodes which execute queries in parallel.