Which of the following provides the MOST cost-effective solution?
Use AWS CloudFormation templates. Deploy a stack with the DB cluster for each development group. Delete the stack at the end of the development cycle.
Use the Aurora DB cloning feature. Deploy a single development and test Aurora DB instance, and create clone instances for the development groups. Delete the clones at the end of the development cycle.
Use Aurora Replicas. From the master automatic pause compute capacity option, create replicas for each development group, and promote each replica to master. Delete the replicas at the end of the development cycle.
Use Aurora Serverless. Restore current Aurora snapshot and deploy to a serverless cluster for each development group. Enable the option to pause the compute capacity on the cluster and set an appropriate timeout.
Explanations:
Using AWS CloudFormation to deploy and delete the DB clusters is not cost-effective. The DB clusters will still incur costs while they are running, even if they are deleted at the end of the cycle. CloudFormation does not provide a mechanism to pause or reduce the cost during inactivity.
Aurora DB cloning creates separate DB clusters for each development group. While it may save on storage, it still requires maintaining multiple instances of Aurora DB clusters, which can lead to high costs due to the need to manage separate DB instances for each group.
Aurora Replicas are used for read scaling and high availability. The solution involves creating replicas and promoting them to master, which can lead to unnecessary costs for maintaining multiple replicas and doesn’t directly address the cost reduction needs when the DB is only used for 8 hours a day.
Aurora Serverless is designed for variable workloads, and it automatically adjusts compute capacity based on demand. The DB can pause during inactivity (e.g., overnight or during non-working hours) and resume when needed, making it the most cost-effective solution. The ability to pause compute capacity when not in use reduces operational costs significantly.
I estimate that the answer is:
Use Aurora Serverless. Restore current Aurora snapshot and deploy to a serverless cluster for each development group. Enable the option to pause the compute capacity on the cluster and set an appropriate timeout.
To the best of my knowledge, the answer is:
Use Aurora Serverless. Restore current Aurora snapshot and deploy to a serverless cluster for each development group. Enable the option to pause the compute capacity on the cluster and set an appropriate timeout.