Cluster_Resize

Add or remove nodes or change node types with resize operation 
There are 3 ways to resize:
1. Elastic resize - Most recommended option - Typically completes in 10 mins or less. You can even schedule this 
2. Classic resize - Upto several hours depending on how much data we have 
3. snapshot and restore with classic resize - If your customer don't want to lose unavailability. 

Classic Resize Under the hood - 
Here, new Redshift cluster of target config is provisioned 
Metadata is copied in 20 mins 
Streaming restore of actual data , takes upto hours 
When restore is completed, CNAME is referenced to new leader node 

Elastic resize Under the hood - 
Slices are migrated to new nodes cluster and hence no. of slices remains same but these slices gets more memory and CPU more. If you want 
you increase the slices then Classic Resize is only the option