Help Center/ Relational Database Service_RDS for SQL Server/ User Guide/ Instance Modifications/ Manually Changing Storage Capacity
Updated on 2026-11-02 GMT+08:00
View PDF
Share

Manually Changing Storage Capacity

Scenarios

If storage is not enough for your increasing workloads, you can scale up the storage for your instance.

If the storage in use is far less than the storage capacity you have requested, scale down the storage capacity to avoid resource waste.

If the storage is full, you need to scale up the storage to make the instance available and resume write operations. You are advised to set alarm rules for the storage space usage by referring to Setting Alarm Rules.

Conditions for an Instance to Become Read-Only When the Storage Is Full

Table 1 Conditions that cause an instance to become read-only due to full storage

Storage Size

Description

Less than 1 TB
  • If the storage usage is greater than or equal to 97%, the instance becomes read-only.
  • If the storage usage is less than or equal to 87%, the instance exits the read-only state.

Greater than or equal to 1 TB
  • If the available storage space is less than 30 GB, the instance becomes read-only.
  • If the available storage space is greater than 150 GB, the instance exits the read-only state.
Figure 1 Storage space usage

How to Disable Read-Only When the Storage Is Full

Locate the instance in the Storage full state and click Disable Read-Only. In the displayed dialog box, click OK.

After disabling read-only mode, scale up the storage in a timely manner. Otherwise, periodic background inspections will set the instance back to read-only and the storage may become completely full, which will interrupt workloads.

Impact on Workloads

For instances with large storage capacities, performing a full backup and re-establishing the primary/standby replication relationship after it has been removed may take a long time. The actual duration depends on the amount of instance data. These operations also have impact on your workload because they consume CPU and disk I/O resources on the backup node.

Constraints

Table 2 Constraints

Phase

Constraints

Duration

Storage scale-up
  • You can scale up storage only when your account balance is greater than or equal to $0 USD.
  • There is no downtime during storage scale-up.
  • The maximum allowed storage is 10,000 GB. There is no limit on the number of scale-ups.
  • The maximum storage capacity for certain RDS for SQL Server instances remains 2,000 GB due to Windows disk size limitations. The actual upper limit is subject to the details displayed on the console.
  • For primary/standby DB instances, scaling up the primary DB instance will cause the standby DB instance to also be scaled up.
  • You cannot reboot or delete a DB instance that is being scaled up.
  • If you scale up the storage of a DB instance with disk encryption enabled, the new storage will still use the original key for encryption.
  • After backup on the standby node is enabled, the maximum capacity for storage scale-up is 32 TB. If storage capacity is expanded to more than 10 TB, backup on the standby node cannot be disabled. For details about how to enable backup on the standby node, see Configuring a Same-Region Backup Policy.

A storage scale-up takes 3 to 5 minutes. Too many databases and usernames may slow down permission replay when an instance exits read-only mode after storage scale-up to resolve insufficient storage. If the instance has more than 100 databases and more than 100 users, the process of disabling read-only status may take several minutes.

Storage scale-down
  • To scale down the storage space of a DB instance, submit a service ticket.
  • The storage space of a DB instance can be scaled only when the instance is Available.
  • Only instances whose storage type is ultra-high I/O, cloud SSD, or extreme SSD can be scaled down.
  • Scaling down storage space of a DB instance during peak hours may fail due to insufficient storage space. Scale down storage space during off-peak hours.
  • You cannot scale down the storage space of a DB instance with publication and subscription enabled.
  • You cannot scale down the storage space of a read replica or a cluster instance with read replicas.
  • Storage scale-down requires that each workload database has at least one automated full backup.
  • Storage scale-down will cause a brief service interruption lasting a few minutes. The interruption duration is directly proportional to the amount of data changed since the last full backup (larger changes result in longer downtime). Evaluate the impact and make necessary arrangements for your workload in advance.
  • When restoring data using backups created before the scale-down, select an instance whose storage capacity is greater than or equal to the storage capacity before the scale-down.
  • To prevent scale-down failures, do not delete or create databases during a scale-down.
  • If storage autoscaling is enabled and the target scale-down capacity is smaller than the threshold for storage autoscaling, storage autoscaling will be triggered automatically after the scale-down is complete. Therefore, select an appropriate target scale-down capacity.
  • After backup on the standby node is enabled, the instance does not support storage scale-down.

Scale down storage during off-peak hours. The time required for scaling down storage depends on the resource utilization and data volume of the DB instance. The larger the data volume, the longer the scale-down process takes.

When there are enough I/O and CPU resources, it takes about 9 hours to scale down a DB instance with 1 TB of data. If the scale-down takes an extended period of time, submit a service ticket.

Procedure

  1. Log in to the RDS console.
  2. Click in the upper left corner and select a region.
  3. On the Instances page, locate the target DB instance and choose More > Change Storage Space in the Operation column.

    Or use either of the following methods:

    • Click the target instance name to enter the Summary page. In the Storage & Backup area, click Change Storage Space.
    • If the instance storage is full, locate the instance on the Instances page and click Scale in the Status column.

  4. On the displayed page, specify the new storage space and click Next.

    You can adjust the storage capacity in increments of at least 10 GB. The selected capacity must be an integer multiple of 10. The instance supports a storage range from 40 GB to 10,000 GB.

    Select a time to schedule the change. (To use this function, submit a service ticket.)

    • Immediate: The change is applied immediately after the request is submitted.
    • During maintenance window: If this option is selected, you can view the task details later on the Scheduled Tasks page. For details, see Viewing a Task. You are advised to set the maintenance window to off-peak hours.

  5. Confirm the information.

    • If you need to modify your settings, click Previous.
    • If you do not need to modify your settings, click Submit for a pay-per-use instance or click Pay Now for a yearly/monthly instance.

  6. Check the result.

    On the Instances page, the status of the DB instance will be Scaling up or Scaling down. After a while, click the instance name and check that the new value for storage space appears on the Summary page.

Parent Topic: Instance Modifications