Configure Auto Vertical Scaling
Auto Vertical Scaling adds compute or storage automatically when usage crosses a threshold, so you are not watching dashboards waiting to resize by hand.
You set the thresholds and the ceiling; the platform does the rest.
Required permissions
| Permission | Action type | Allows |
|---|---|---|
manageDatabase:Configuration | Edit | Modifying cluster configuration, including scaling settings |
Prerequisite: notifications first
The Notification service must be enabled with at least one email or Telegram recipient before scaling can be turned on. Without somewhere to send scaling alerts, the console blocks activation and offers a Go to Notification link.
Set that up first — see Notification.
Enable the service
You can enable Auto Vertical Scaling during provisioning, in the additional services step of Create your first database — which is the simpler route, since the prerequisite is handled in the same wizard.
To enable it on an existing cluster:
- From the main menu, select Database Platform → All Databases or Non-Relational Database. Click the Cluster ID and open the Vertical Scaling tab. A cluster without the service shows Inactive.
- Click Enable Vertical Scaling — the gear icon next to Service Information.
- Click Confirm.
The status changes to Active.
Configure compute scaling
On the Vertical Scaling page, click Add Flavor Scaling (plus icon) under Compute Scaling.
| Field | What it controls |
|---|---|
| CPU Threshold (%) | CPU usage that triggers scaling. At 60%, scaling starts once CPU passes 60%. |
| RAM Threshold (%) | Memory usage that triggers scaling. |
| Tracing Interval | How often usage is evaluated, in minutes. |
| Current Flavor | The node's present configuration, for reference. |
| List Flavor Scaling | The flavors the system may scale up to, used in order. |
How the flavor list works: with a current flavor of Medium-4 and a list of Medium-8 then Large-2, crossing the threshold scales to Medium-8. If usage is still above the threshold afterwards, it scales again to Large-2. The list is the ceiling — scaling stops at the last entry.
Click Add to save. Edit later with the pencil icon, or remove the configuration with the trash icon to stop automatic scaling.
- Quota is the hard ceiling. Scaling happens only within your available quota; beyond it, scale-up is refused regardless of the flavor list.
- Do not set thresholds too low. Below 60% invites continuous scaling that wastes resources without improving anything.
- Watch CPU and RAM trends so you can request quota before scaling hits the ceiling.
Configure storage scaling
On the Vertical Scaling page, click Add Storage Scaling (plus icon) under Storage Scaling.
| Field | What it controls |
|---|---|
| Storage Threshold (%) | Usage that triggers expansion. |
| Storage Scale (%) | How much capacity is added per operation, as a percentage of current total. |
At a 20% storage scale, a 100 GB disk becomes 120 GB when the threshold is crossed. Storage scaling runs online, without service disruption.
Click Add to save.
Storage scaling is one-way. Once capacity is added it cannot be reduced, and you keep paying for it. Set the threshold high enough (70% or above) that you are not expanding on transient spikes.
Next steps
- Change database resource configurations for manual adjustments
- Notification to control who hears about scaling events