What is Redshift?
Redshift is a column-based data warehouse. So, what is a columnar database, you might ask? Most of the databases you will come across would be considered row-based databases. These architectures are common in transactional databases, also known as Online Transaction Processing (OLTP), which provide the backend for most enterprise applications. When a query is sent to a row-based database, it will read through its table one row at a time, reading all the attributes from all of the columns in that row, before moving on to the next row of data. These databases are good at reading tables that may have many columns and fewer rows. When I say fewer rows, I typically mean under a billion, but that exact number can depend on a few variables.
Where columnar databases differ is that when they read a table, they read it column by column. They first read every value within the first column before reading data from the second column. So, if I want to only read the first column...