Database Autonomy Service (DAS) monitors average bandwidth usage in real time and automatically scales bandwidth up or down for Redis instances—so you can handle traffic spikes without manual intervention.
Elastic capabilities typically have a lag of 3 to 5 minutes. For latency-sensitive workloads, lower the Auto Scaling thresholds and reduce the monitoring frequency.
Auto Scaling becomes unavailable when your account has an overdue payment. The feature resumes within 3 hours after your account status returns to normal.
Automatic bandwidth adjustment relies on real-time performance trend data. If this data is interrupted or significantly delayed, the feature stops operating.
Supported instance types
Automatic bandwidth adjustment is available for instances using the standard, cluster, or read/write splitting architecture of the following:
Redis Open-Source Edition (classic or cloud-native)
Memory-optimized Tair (Redis OSS-compatible) (classic or cloud-native)
Persistent memory-optimized Tair (Redis OSS-compatible)
A DAS service-linked role must exist for your account. If not, DAS creates one automatically when you first enable automatic bandwidth adjustment.
Use cases
| Scenario | Description |
|---|---|
| Handle traffic spikes | Automatically increase bandwidth during flash sales or promotional events, then decrease it afterward to reduce costs. |
| Mitigate large-key impact | When heavy read/write operations on large keys strain bandwidth, temporarily increase it to keep your service stable while you address the root cause. |
| Reduce request skew costs | In cluster or read/write splitting architectures, some data shards or read replicas get more traffic than others. Auto Scaling identifies the overloaded ones and increases their bandwidth—no manual spec changes needed. |
How it works
When you enable bandwidth auto scaling, the system monitors your instance based on the observation window and thresholds you configure. Bandwidth increments and decrements are calculated automatically:
Scale up: When average bandwidth usage reaches or exceeds the increase threshold, the system increases bandwidth. If the threshold is reached again, the system increases bandwidth again. The maximum increase is 6× the default bandwidth of the instance type, capped at an additional 192 MB/s.
Scale down: When average bandwidth usage falls to or below the decrease threshold, the system decreases bandwidth. If the threshold is reached again, the system decreases bandwidth again. Bandwidth cannot go below the default bandwidth of the instance type.
Target bandwidth calculation
The target bandwidth after each scaling operation is:
Target bandwidth (MB/s) = Actual bandwidth usage (MB/s) / ((Increase threshold + Decrease threshold) / 2)After each operation, the system keeps actual bandwidth usage between the two thresholds.
Scale-up example: An instance has a default bandwidth of 96 MB/s. The increase threshold is 70%, the decrease threshold is 30%, and the observation window is 15 minutes. When average bandwidth usage reaches 70%, the system calculates the target bandwidth as:
(96 MB/s × 70%) / ((70% + 30%) / 2) = 135 MB/sScale-down example: Using the same instance, if average bandwidth usage falls to or below 30% after the scale-up, the system scales down. The minimum bandwidth is 96 MB/s (the default for this instance type).
For instances that need more bandwidth than the 6× cap allows, consider upgrading to Tair (Enterprise Edition), which supports a minimum of 96 MB/s per instance type. See Instance types for details.
Limitations
Cooling periods: After an automatic bandwidth increase, the system waits at least 1 hour before triggering an automatic decrease. Between two consecutive automatic increases, there is a 1-minute cooling period.
Maximum bandwidth cap:
Bandwidth can increase by up to 6× the default bandwidth, with a maximum additional 192 MB/s.
Example — 2 GB standard Tair memory-optimized instance: default 96 MB/s, maximum 288 MB/s (96 + 192).
Example — 256 MB standard Redis Community Edition instance: default 10 MB/s, maximum 70 MB/s (10 + 60, because the 6× cap equals 60 MB/s).
To exceed these limits, upgrade the instance type or architecture (for example, from standard to cluster). See Instance types.
Unexpired bandwidth plans: If the instance has unexpired bandwidth plans, unsubscribe from those plans before enabling auto scaling. See Unsubscription management.
Operations that disable auto scaling: The following operations disable bandwidth auto scaling. Re-enable it after the operation completes if needed.
| Operation | Exception for standard architecture |
|---|---|
| Major version upgrade | None |
| Change instance configurations | Bandwidth settings remain valid after a spec change. |
| Change the zone of an instance | Bandwidth settings remain valid after zone migration. |
Considerations
Avoid mixing manual and auto bandwidth adjustment. If you set bandwidth manually and also enable auto scaling, auto scaling stops triggering increases once the manually set bandwidth hits the 6× cap. As traffic decreases, auto scaling reduces bandwidth back to the default.
Example: Default bandwidth is 10 MB/s, manually set to 70 MB/s. Auto scaling cannot increase further because 70 MB/s already equals the 6× cap (60 MB/s additional). Auto scaling can still decrease bandwidth toward the 10 MB/s default.
Example: Default bandwidth is 10 MB/s, manually set to 40 MB/s. Auto scaling can increase up to 70 MB/s total, and decreases toward the 10 MB/s default when the decrease threshold is met.
Compare to instance type changes. Bandwidth adjustment is faster, costs less overall, and avoids transient disconnections. To adjust bandwidth manually instead, see Manually adjust bandwidth for Redis instances.
Billing
Bandwidth beyond the default is billed hourly based on the amount of additional bandwidth and its usage duration. Rates vary by region. See Billing items for pricing details.
The default instance bandwidth is free. Charges apply only to bandwidth added beyond the default.
Enable automatic bandwidth adjustment
Log on to the DAS console.
In the left navigation pane, click Intelligent O&M Center > Instance Monitoring.
Find the target instance and click the instance ID to open the instance details page.
In the left navigation pane of the instance page, click Autonomy Center. On the right side of the page, click Autonomy Feature Switch.
On the Autonomous Function Management > Autonomous Function Settings tab, turn on the autonomy feature switch.
On the Optimization and Rate Limiting tab, configure automatic bandwidth adjustment:
Category Parameter Description Automatic bandwidth upgrade Automatic bandwidth upgrade Select to enable. Average bandwidth usage not less than The increase threshold (%). Auto scaling triggers when average bandwidth usage meets or exceeds this value. The system uses the higher of inbound or outbound usage. Observation window The monitoring window duration in minutes. Automatic bandwidth downgrade Automatic bandwidth downgrade Select to enable. Requires Automatic bandwidth upgrade to be enabled first. Average bandwidth usage not greater than The decrease threshold (%). Auto scaling triggers when average bandwidth usage falls to or below this value. The system uses the lower of inbound or outbound usage. Click OK.
Optional: Click Event Subscription Settings to configure notifications for automatic bandwidth upgrades or downgrades.
When an automatic bandwidth change occurs, DAS generates a notification-level event. Turn on Enable subscription service and configure event notification parameters. For more information, see Enable event subscription.
(Optional) Click Event Subscription Settings to configure notifications for bandwidth changes. When an automatic bandwidth change occurs, DAS generates a notification-level event. Turn on Enable subscription service and configure the event notification parameters. See Enable event subscription.
(Optional) Configure an alert template to receive alerts about bandwidth changes. The system recommends a template and adds autonomy event alert rules automatically. Follow the prompts to complete the configuration.
NoteIf the instance already has an alert template, follow the prompts to add autonomy event alert rules to it.
To create a custom template and rules, see Configure alert templates and Configure alert rules.
Select an alert contact group to receive notifications: See Manage alert contacts for detailed steps.
Click Add contact to add a new alert contact.
Click Create contact group to add a new contact group.
Click Edit or Remove next to a contact to update or delete their information.
Click Submit configuration, then confirm the alert settings in the dialog box.