First setup of Redshift serverless

0

Hello,

I am totally new when it comes to Redshift so I have a question about the setup. What is the best practice when setting up Redshift when we have a AWS DEV, UAT and PROD environment? Should I create 3 instances of Redshift or 1 is sufficient and work with namespaces/workgroups?

If the namespace/workgroup solution will work, do 1 use 1 namespace with different workgroups(dev, uat and prod) or something else?

Thanks for the help

René

Rene
asked 7 months ago214 views
1 Answer
0
Accepted Answer

This really depends on how big/small your organization is, what kind of SDLC policies are in place for your application development, what domain the company belongs to, example, financial services domain might follow different regulations compared to medical sciences or compared to federal organizations. If I had a free hand to build out a new use case then I would go as follows...

If you are exploring a use case then you can start with one workgroup and namespace. Create your tables, load some data, run your queries and validate this works for you your use case. This can be treated as you Development setup as you are freely making changes as you see fit. At this stage most likely you are the only producer and consumer of this data. So this can be considered as Production but only for you.

Once you have something concrete then you might want more control over changes, you might also have a team of resources working on this project at the same time. You might want to validate any changes that go in new do not break anything. Thats when you can replicate your Dev setup into a new AWS account, make the QA team the owner of that, and this is your QA setup.

And finally when you have real data feeds coming in, you have real consumers accessing your data and depending on it to make business critical decisions then it becomes a Production use case. So you might want to setup another environment in another AWS account for Production usage.

Typically I have seen different data volumes, different number of users, different number of queries running on Production Vs. QA Vs. Development. For example, Production could be used continuously and executing a repeated set of queries, QA running infrequently - only when a new release is going to be tested, where as Development might be used say 9-to-5 and having a wide variety of queries executing. Having different AWS accounts with different workgroups + namespace allows you to implement controls and also size each environment depending on the needs.

profile pictureAWS
answered 7 months ago
  • Thanks for this answer, definitely worth to consider!

You are not logged in. Log in to post an answer.

A good answer clearly answers the question and provides constructive feedback and encourages professional growth in the question asker.

Guidelines for Answering Questions