Amazon EC2 allows you to stop your EBS-based EC2 instances and restart them. While the instance is stopped, you’re not paying for the instance (with the exception of your EBS storage). If your EC2 instance only needs to be running during parts of the day, this ability to stop and restart the instance allows you to save some money during times that your EC2 instance is not needed.
Unfortunately, Amazon RDS does not have the same capabilities to stop and restart the instance. Instead, the only controls you have of your RDS instance is to restore a new instance from a previous snapshot, and to delete an RDS instance, creating a final snapshot in the process.
However, these minimal controls still allow you to achieve similar cost-savings with your RDS instance as you have with your EC2 instances. Skeddly allows you to use these building-blocks to save money on your RDS costs by restoring your RDS instance when it’s needed, and deleting it when it’s not.
Let’s say that your RDS instance is only needed from 8am to 8pm, Monday through Friday. You can achieve cost savings using Skeddly by following these steps:
Below is a screenshot of these settings in Skeddly.
When the action executes it will create a fresh RDS instance based on the previous night’s final snapshot.
By using Skeddly to implement a 12-hour, Monday through Friday up-time schedule, you can realize a cost-savings of 64%. For a db.m3.medium MySQL instance, that’s a savings of $506 per year. For larger instance types, the cost-savings grows even higher, such as with a db.r3.8xlarge where the cost-savings could be over $42,000 per year.