Changing RDS storage from gp2 to gp3

Here are few pointers, that you need to keep in mind:

1. General Purpose SSD gp3 storage is supported on Single-AZ and Multi-AZ DB instances, but isn't supported on Multi-AZ DB clusters.
2. For every DB engine except RDS for SQL Server, you can provision additional IOPS and storage throughput when storage size is at or above the threshold value. For RDS for SQL Server, you can provision additional IOPS and storage throughput for any available storage size. For all DB engines, you pay for only the additional provisioned storage performance.
3. There is no downtime associated but always plan to make modifications at the time when clusters is expected to be least busy. There wouldn't be downtime, you may see “Storage Optimization” status.
4. GP2 to GP3 is no brainer.

You can make the changes through console or through CLI:

    aws rds modify-db-instance \
--db-instance-identifier your-instance-name \
--storage-type gp3 \
--apply-immediately

Here are some artifacts for your reference:

Key highlights of GP3 storage volumes Comparison of RDS storage options

Apart from above, I would still consider backup restore operation to be clear and in place as due diligence and best practice.

Everything fails, all the time

When reviewing the documentation here: https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/Overview.DBInstance.Modifying.html#USER_ModifyInstance.Settings and then scrolling down to the section "Storage type" you can see that there is "brief downtime while the process starts" for changes:

• From General Purpose (SSD) or Provisioned IOPS (SSD) to Magnetic.
• From Magnetic to General Purpose (SSD) or Provisioned IOPS (SSD).

Note that this implies switching from gp2 (SSD) to gp3 (SSD) will not result in downtime but it does not state this explicity.

Reminders: you should not accept answers to specific questions unless they are vendor-supplied or a conclusive test with data has been provided. You can see above that every single answer previously submitted was wrong.

Changing storage type from GP2 to GP3 has no downtime or reboot/failover. The only thing is the DB will then be in Storage Optimization but the instance will be accessible and you can perform regular operations.

This document explains more about the storage optimization state --> https://repost.aws/knowledge-center/rds-stuck-in-storage-optimization

This process should not introduce any risks, gp3 is a more performant and more cost effective EBS storage type. Just you should make sure that there is a proper change management as there will be an outage. As far as you are taking the snapshot, you should be good.

Hello,

Thank you for asking your question here.

AWS RDS SQL Server is a managed service and it is designed to take care of any disaster. Please go ahead and modify the storage, our internal automation will take care of data integrity.

Storage optimization will occur for any type of RDS storage modification.

https://repost.aws/knowledge-center/rds-stuck-in-storage-optimization