I'm struggling to understand what the difference is between a user group and a query group when manually implementing query queues in AWS Redshift. What is the difference in implementation?
These are 2 properties that can be used in deciding query placement into a queue. One, user group, has lots of meaning within Redshift like assigning user privileges and can only be changed by an admin as it is a database-wide configuration. The other, query group, is a user settable (session) property that isn't overloaded with additional meaning. User group = static and global. Query group = dynamic and connection specific.
These 2 properties are used by the workload manager as inputs when choosing queues for queries. I have found the graphic on this page to be very helpful in understanding the WLM decision process. https://docs.aws.amazon.com/redshift/latest/dg/cm-c-wlm-queue-assignment-rules.html