A data analytics company has an Amazon Redshift cluster that consists of several reserved nodes.

The best solution is to deploy an AWS Lambda function to add capacity to the Amazon Redshift cluster by using an elastic resize operation when the cluster’s CPU metrics in Amazon CloudWatch reach 80%. This solution will enable the cluster to scale up or down quickly by adding or removing nodes within minutes. This will improve the performance of the complex read queries and also reduce the cost by scaling down when the demand decreases. This solution is more cost-effective than using a classic resize operation, which takes longer and requires more downtime. It is also more suitable than using Amazon EMR, which is designed for big data processing rather than data warehousing. References: Amazon Redshift Documentation, Resizing clusters in Amazon Redshift, [Amazon EMR Documentation]