Change the configurations of an ApsaraDB RDS for MySQL instance — including the instance edition, instance type, and storage — to match your evolving workload requirements.
For a full list of configurable items, see Instance configuration changes.
To change configurations for instances running other database engines, see:
-
Change the configurations of an ApsaraDB RDS for SQL Server instance
-
Change the configurations of an ApsaraDB RDS for PostgreSQL instance
-
Change the configurations of an ApsaraDB RDS for MariaDB instance
Prerequisites
Before you begin, ensure that:
-
The instance uses the subscription or pay-as-you-go billing method. For serverless instances, see Configure a serverless ApsaraDB RDS for MySQL instance.
-
Your Alibaba Cloud account has no overdue renewal payments.
-
The instance is in the Running state. If it is locked, unlock the instance first.
Limitations
Order
After you submit a configuration change order, you cannot cancel it. Evaluate your business requirements carefully before proceeding.
Read-only instances
-
The primary instance must be in the Running state when you change a read-only instance.
-
The storage of a read-only instance must be greater than or equal to the current storage of its primary instance. Expand the storage of all read-only instances before expanding the primary instance.
Storage scale-in
Local disks with enhanced performance:
-
Storage after scale-in must be at least 120% of the used storage.
-
Example: If an instance has 100 GB total with 50 GB used, the minimum post-scale-in storage is 60 GB (50 × 120%).
General limits (Basic Edition and High-availability Edition instances of the same edition and architecture):
-
Minimum storage after scale-in:
min{current usage × 1.3, current usage + 400 GB}, and no less than the minimum storage supported by the current instance type. -
Step size: 5 GB increments.
-
If binary logs are generated rapidly, retain sufficient logs on the local disk before scaling in. For details, see Modify the backup policy for an ApsaraDB RDS for MySQL instance.
Product type changes
-
Zone compatibility: Yitian Edition supports fewer zones than Standard Edition. If zone incompatibility blocks a product type change, first change the zone, then change the product type.
-
Minor engine version: The target minor engine version must be the same as or later than the current version. Downgrading the minor engine version is not supported.
-
Edition constraints:
-
Basic Edition: can only change from Yitian Edition to Standard Edition.
-
High-availability Edition and Cluster Edition: can change between Yitian Edition and Standard Edition.
-
Legacy instance types
Legacy instance types cannot be changed directly. First upgrade the instance to a current instance type, then change its configurations.
Potential impacts
Configuration changes involve two phases:
-
Data migration (instance stays online): A new instance is provisioned with the target configuration. Data is synchronized to the new instance. Your current instance remains online and accepts connections throughout this phase.
-
Switchover (brief interruption): At the switchover time, the original instance ID and endpoint are redirected to the new instance. The switchover typically takes about 30 seconds. The actual duration may be longer if the client is not properly configured or the driver version is earlier than the required version. During this window, most operations related to databases, accounts, and networks cannot be performed. For details on the full impact, see Impacts of an instance switchover.
The table below describes what to expect for each change type.
| Storage type | Change item | Impact |
|---|---|---|
| Local disks with enhanced performance | Instance type, edition, or storage | If you change the instance type or edition when no local resources are available, data is automatically migrated. After the migration is complete, a switchover is performed based on the switchover time that you specify. During the migration, incremental data synchronization is performed. |
| Instances that use disks | Instance type or edition | Takes several minutes, regardless of data volume. A switchover occurs. |
| Instances that use disks | Storage scale-out | No interruption for ESSDs or premium performance disks. A transient connection occurs for standard SSDs. |
| Instances that use disks | Storage scale-in | A transient connection occurs. See Scale in the storage space of an instance that uses disks. |
If a snapshot is being created for a disk, storage expansion or performance level changes are queued until the snapshot completes.
After the change completes:
-
No manual restart is required, and no stored data is lost.
-
The instance ID and endpoint remain unchanged. If the instance is migrated across hosts, the IP address behind the endpoint may change. Always connect using the endpoint, not the IP address.
Schedule configuration changes during off-peak hours. Make sure your application has an automatic reconnection mechanism. Basic Edition instances have a single database node with no secondary node for hot backup — a brief service interruption occurs during the change.
Billing
For more information, see Billing rules for configuration changes.
Change the configurations
After you submit a configuration change order, you cannot cancel it. Evaluate your business requirements carefully before proceeding.
-
Go to the Instances page. In the top navigation bar, select the region where the instance resides. Find the instance and click its ID.
-
In the Configuration Information area, click Change Specifications.
-
(Subscription instances only) In the dialog box, select a change method and click Next. After submission, the system synchronizes data to a new instance. At the switchover time, the original instance ID and endpoint are redirected to the new instance.
Method Description Upgrade Increases specifications. Takes effect immediately after the change. Downgrade Decreases specifications. Takes effect immediately after the change. -
On the Upgrade/Downgrade Instance page, update the instance configurations.
Note-
All instance types support instance type changes and storage adjustments. For legacy instance types, upgrade to a current instance type before expanding storage.
-
Storage scale-in is supported for High-availability Edition instances that use local disks with enhanced performance, and for Basic Edition or High-availability Edition instances that use disks (excluding standard SSDs).
-
You can change the edition of a three-node ApsaraDB RDS for MySQL 5.6 Enterprise Edition instance to High-availability Edition.
-
MySQL 5.7 Basic Edition instances can be upgraded to High-availability Edition with local disks with enhanced performance.
-
MySQL 8.0 or 5.7 High-availability Edition instances using ESSDs can be changed to Cluster Edition.
-
The instance type defaults to Current Type on the upgrade/downgrade page. Verify that the target instance type meets your requirements before proceeding.
-
-
Under Switch Time, select when the switchover occurs:
Important- If you select Switch Within Maintenance Window, the instance stays in the Upgrading or Downgrading state until the switchover completes. During this period, instance-level operations such as version upgrades and cross-zone migrations are unavailable. - If you are only expanding storage or changing the ESSD performance level, the change takes effect immediately with minimal service impact. Selecting Switch Within Maintenance Window is not necessary. - For non-Basic Edition instances, one or two switchovers may occur, causing a transient connection. Perform changes during off-peak hours, or make sure your application has an automatic reconnection mechanism. - Basic Edition instances have only one database node with no secondary node for hot backup. A service interruption occurs when you change the configurations. Change the configurations during off-peak hours to prevent business interruptions.
Option Behavior Execute Immediately Switchover occurs immediately after data migration completes. Switch Within Maintenance Window Switchover occurs during the next maintenance window. -
Review the changes, click Confirm Order, and complete the payment.
Warning- After you submit the order, you cannot cancel it. - Do not run Data Definition Language (DDL) operations from the time you submit the order until the configuration change completes.
To monitor the progress of the change, click the task list button in the upper-right corner of the Basic Information page to open the Task List.
FAQ
Disk scale-out and scale-in
Why is my instance still locked after I expanded the disk?
An instance locked due to a full disk is automatically unlocked once the scale-out task completes. To check progress, click the 
button in the upper-right corner of the Basic Information page to open the Task List.
Why does a transient connection occur when I expand storage?
Storage expansion triggers a switchover. For details on the connection impact, see Impacts of an instance switchover.
Does the free backup quota increase after I expand the disk?
Yes. For details, see Free backup quota.
Storage type changes
How do I change the storage type (local disk with enhanced performance, standard SSD, or ESSD)?
See Change the storage type of an instance that uses disks to local disks with enhanced performance.
How do I enable automatic storage expansion for an instance that uses local disks with enhanced performance?
Change the storage type to a disk-based type and then enable automatic storage expansion. Alternatively, create a new disk-based instance and migrate data to it.
The storage of my instance with local disks with enhanced performance has reached its limit. How do I increase it?
Change the storage type to a disk-based type to access larger storage capacity. For the storage range of MySQL 8.0 or 5.7 High-availability Edition instances with local disks, see ApsaraDB RDS for MySQL High-availability Edition (with local disks).
Storage space management
What do I do if the "Insufficient Inventory" message appears when I expand storage?
Migrate the instance to another zone and then expand the storage. The endpoint remains unchanged after migration, but the IP address changes — configure an automatic reconnection mechanism in your application.
If I only expand storage, is data migrated to a new instance?
For disk-based instances (excluding Basic Edition), storage expansion generally does not cause a transient connection.
For instances with local disks with enhanced performance:
-
Host has sufficient storage: Storage expands directly with no instance impact.
-
Host has insufficient storage: A new primary and secondary instance are created and data is synchronized. A transient connection lasting about 30 seconds occurs during the switchover.
The remaining storage on the host is not visible in the console.
If I delete a database from an RDS instance, is the disk space released?
The DROP statement releases disk space. The DELETE statement causes fragmentation and does not release space.
Impacts of configuration changes
If I upgrade CPU, memory, and disk at the same time, how long is the service unavailable?
The service is unavailable for several minutes, whether you change one item or multiple items at once. An instance switchover or restart may occur, and most operations related to databases, accounts, and networks are unavailable during this period. Perform changes during the maintenance window to minimize business impact. For per-change impact details, see the Potential impacts section.
What factors affect how long a configuration change takes?
See What factors affect the duration of a configuration change for an ApsaraDB RDS for MySQL instance?
Read-only instances
If I upgrade the primary instance, are read-only instances upgraded automatically?
No. Upgrade each read-only instance separately.
Does expanding the primary instance's storage affect replication to read-only instances?
No.
Connections and network
Does the endpoint change after a configuration change?
No. The endpoint (for example, rm-bpxxxxx.mysql.rds.aliyuncs.com) remains unchanged. The IP address behind it may change, so always use the endpoint in your application rather than the IP address.
How do I configure an automatic reconnection mechanism?
For Java applications, set the DNS Time to Live (TTL) to 60 seconds or less. This makes sure the application re-queries DNS to get the updated IP address when the endpoint's IP changes. For TTL configuration in Java, see the official JDK documentation.
Instance types and billing
Can I change a regular instance to a serverless instance?
A: No. To use a serverless instance, you must create one. For more information, see Create a serverless instance.
Subscription instances cannot be converted to serverless instances. To use serverless, create a new serverless instance. See Create a serverless ApsaraDB RDS for MySQL instance.
How am I charged for automatic storage expansion?
Automatic storage expansion is disabled by default and is free to enable. When triggered, charges apply using the same rules as manual expansion. See Billing rules for configuration changes.
I accidentally downgraded and then upgraded back to the original configuration. Why is the price difference so large?
The instance may have been purchased during a promotional period. If you upgrade after the promotion ends, the standard price applies with no discount.
Automatic expansion and scale-in
What are the rules for automatic storage expansion?
See Enable automatic storage expansion.
What are the impacts of scaling in storage on a disk-based instance?
A transient connection occurs. See Scale in the storage space of an instance that uses disks.
Can I scale in an instance that uses local disks with enhanced performance?
Yes. While the instance is running, click Change Configuration and select Downgrade.
Zone migration
How do I change from a single-zone deployment to a multi-zone deployment?
API reference
| API | Description |
|---|---|
| ModifyDBInstanceSpec | Changes the configurations of an ApsaraDB RDS instance. |