1 Answer
- Newest
- Most votes
- Most comments
1
As per Redshift best practices for well-architected frame work, two nodes are the recommended approach for business-critical or production environments. When you have a single node Ra3 cluster, your leader node and compute node are the same, and any issue with a node may require additional downtime.
1 node, RA3. Xlpus also has some other limitations, like only classic resizing and the limited total number of total tables supported. If you are concerned with cost, you may want to look into
- Reserved instances in Redshift cost between 30 and 60% less than on-demand instances. a. AWS-Pricing-calculator
- Based on the usage pattern, Redshift Serverless may be cost-efficient as it will auto-pause. a. https://aws.amazon.com/redshift/redshift-serverless/
Relevant content
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated a year ago
- AWS OFFICIALUpdated 2 years ago
- AWS OFFICIALUpdated 2 years ago