Redshift is a fast, fully managed, petabyte scale, SQL based, data warehouse for large volumes of aggregated data.

It is optimised for OLAP ( online analytics processing ) and allows to perform complex queries on massive collection of data that is structured or semi-structured.

It provides continuous incremental backups and always keeps 3 copies of your data.

Since it uses EC2 instances, you will pay based on the instance family/type.

Data can be sourced from

DynamoDB

RDS

S3

EMR (Elastic Map Reduce)

Data Pipeline

Glue

EC2

On-premises servers

Redshift

A cluster is the main component of Redshift and each cluster will run its own Redshift engine, with at least one Database.

A Cluster is a logical grouping of nodes. When more nodes are present, a Leader Node has the role of coordinating communication between you nodes and external applications accessing your data.