Amazon RDS Storage Can Now Be Modified Up to 4 Times Within 24 Hours, So I Tried It Out

Amazon RDS Storage Can Now Be Modified Up to 4 Times Within 24 Hours, So I Tried It Out

Amazon RDS now allows up to 4 storage size, type, or performance changes per 24 hours, removing the old 6-hour cooldown. On RDS for PostgreSQL, 4 consecutive size changes succeeded and the 5th returned an error.
2026.07.16

This page has been translated by machine translation. View original

Introduction

With the update on July 15, 2026, Amazon RDS now allows storage size changes, storage type changes, and performance (IOPS/throughput) changes to be performed up to 4 times per 24 hours. The 6-hour cooldown that was previously required after every storage modification has been removed.

https://aws.amazon.com/about-aws/whats-new/2026/07/amazon-rds-upto-four-storage-modifications-24-hours

Item Old specification New specification (2026-07-15 onward)
Cooldown after storage modification 6-hour wait required Next modification possible immediately once optimization completes
Maximum modifications per 24 hours Effectively 4 (but a 6-hour wait was required after each modification) 4 (can be run consecutively as soon as optimization completes)
Modification during storage-optimization Not allowed Not allowed (no change)

A similar relaxation was applied earlier on the Amazon EBS side in January 2026. This time, the same limit relaxation that EBS had adopted earlier has also been introduced for RDS storage modifications. The verification on the EBS side is summarized in the following article.

https://dev.classmethod.jp/articles/ebs-elastic-volumes-4-modifications/

In this article, I try consecutive storage size modifications on RDS for PostgreSQL and confirm that 4 modifications succeed and that the 5th results in an error. Storage type changes and IOPS changes have already been verified in the EBS (EC2) article, so they are out of scope for this article.

Verification

Verification environment

Item Value
AWS CLI 2.35.22
Region ap-northeast-1
Instance identifier test-storage-4mod
Engine PostgreSQL 18.3
Instance class db.t4g.micro
Storage type gp3
Initial storage 20 GB
Initial IOPS 3000
Initial throughput 125 MB/s
Multi-AZ No (Single-AZ)
Public access No
Deletion protection No
Automated backup retention period 1 day

Creating the RDS instance

I created an RDS for PostgreSQL instance for verification with the following command.

aws rds create-db-instance \
    --db-instance-identifier test-storage-4mod \
    --db-instance-class db.t4g.micro \
    --engine postgres \
    --engine-version 18.3 \
    --allocated-storage 20 \
    --storage-type gp3 \
    --iops 3000 \
    --storage-throughput 125 \
    --master-username postgres \
    --manage-master-user-password \
    --no-multi-az \
    --no-publicly-accessible \
    --no-deletion-protection \
    --backup-retention-period 1 \
    --region ap-northeast-1

It became available in about 6 minutes.

aws rds wait db-instance-available --db-instance-identifier test-storage-4mod --region ap-northeast-1

Checking the gp3 storage size constraint

Before the consecutive modifications, I check the modifiable storage sizes with describe-valid-db-instance-modifications.

aws rds describe-valid-db-instance-modifications \
    --db-instance-identifier test-storage-4mod \
    --region ap-northeast-1

The StorageSize included in ValidStorageOptions in the response was as follows.

"StorageSize": [
    {
        "From": 22,
        "To": 65536,
        "Step": 1
    }
]

With an initial size of 20 GB, the minimum value that could be specified for the next modification was 22 GB. For gp3 storage size modifications, there is a constraint that you must "increase by at least 10% from the current size." I had originally planned to change in 1 GB increments, but this constraint made that impossible. The valid minimum sizes confirmed with describe-valid-db-instance-modifications each time are summarized below.

Current size Next valid minimum size Increase
20 GB 22 GB +2 GB (10%)
22 GB 25 GB +3 GB (≈14%)
25 GB 28 GB +3 GB (12%)
28 GB 31 GB +3 GB (≈11%)
31 GB 35 GB +4 GB (≈13%)

1st: 20 GB → 22 GB

aws rds modify-db-instance \
    --db-instance-identifier test-storage-4mod \
    --allocated-storage 22 \
    --apply-immediately \
    --region ap-northeast-1

PendingModifiedValues.AllocatedStorage in the response was 22. I checked the status with the following command.

aws rds describe-db-instances \
    --db-instance-identifier test-storage-4mod \
    --query 'DBInstances[0].DBInstanceStatus' \
    --output text \
    --region ap-northeast-1
  • Status transition: available → modifying → storage-optimization → available
  • Duration of storage-optimization: about 10 minutes

2nd: 22 GB → 25 GB

aws rds modify-db-instance \
    --db-instance-identifier test-storage-4mod \
    --allocated-storage 25 \
    --apply-immediately \
    --region ap-northeast-1

PendingModifiedValues.AllocatedStorage in the response was 25. I checked the status in the same way as the 1st time.

  • Status transition: available → modifying → storage-optimization → available
  • Duration of storage-optimization: about 15 minutes

3rd: 25 GB → 28 GB

aws rds modify-db-instance \
    --db-instance-identifier test-storage-4mod \
    --allocated-storage 28 \
    --apply-immediately \
    --region ap-northeast-1

PendingModifiedValues.AllocatedStorage in the response was 28. I checked the status in the same way.

  • Status transition: available → modifying → storage-optimization → available
  • Duration of storage-optimization: about 20 minutes

4th: 28 GB → 31 GB

aws rds modify-db-instance \
    --db-instance-identifier test-storage-4mod \
    --allocated-storage 31 \
    --apply-immediately \
    --region ap-northeast-1

PendingModifiedValues.AllocatedStorage in the response was 31. I checked the status in the same way.

  • Status transition: available → modifying → storage-optimization → available
  • Duration of storage-optimization: about 20 minutes

5th: 31 GB → 35 GB (exceeding the 24-hour limit)

With the 4th optimization complete and the instance back to available, I attempted the 5th modification.

aws rds modify-db-instance \
    --db-instance-identifier test-storage-4mod \
    --allocated-storage 35 \
    --apply-immediately \
    --region ap-northeast-1

It resulted in an InvalidParameterCombination error.

An error occurred (InvalidParameterCombination) when calling the ModifyDBInstance operation: You can't modify the storage because you've exceeded the maximum number of storage modifications allowed in the last 24 hours. Try again in approximately 22 hours.

Supplement: Attempting a modification during storage-optimization

Attempting the next modification in the middle of storage-optimization also results in an InvalidParameterCombination error.

An error occurred (InvalidParameterCombination) when calling the ModifyDBInstance operation: You can't currently modify the storage of this DB instance because the previous storage change is being optimized.

Cleanup

After the verification, I deleted the instance. Since this was for verification purposes, I deleted it without keeping a final snapshot or automated backups.

aws rds delete-db-instance \
    --db-instance-identifier test-storage-4mod \
    --skip-final-snapshot \
    --delete-automated-backups \
    --region ap-northeast-1

Summary

We confirmed that the relaxation of storage modifications to "up to 4 times per 24 hours," which EBS had adopted earlier, has also been applied to RDS. Please make use of it for things like incrementally scaling up production environments or exploring the optimal storage size in load-testing environments.

Share this article

AWSのお困り事はクラスメソッドへ