Skip to main content

Time-Based Scaling

Time-based scaling changes a cluster's compute capacity at fixed times each day. Use it for predictable workload peaks and low periods, such as regular business hours and overnight traffic.

A policy requires at least two rules with different Target vCPU values. When a rule scales out compute, cache capacity increases in proportion to compute capacity. When a rule scales in compute, cache capacity decreases to the greater of the proportional value and the minimum cache capacity for the target vCPU.

Warning:

  • During scaling, some connections might disconnect and reconnect intermittently.
  • When scaling in reduces cache capacity, the oldest cached data beyond the new capacity is deleted, and response times for some requests might increase.
  • Scaling out compute can increase compute charges.

Rule requirements and behavior​

  • Rules are scheduled to run every day. Execution Period is fixed to Every day.
  • Each rule must have a unique Execution Time.
  • Keep at least one hour between rule execution times.
  • You can configure up to 24 rules for a cluster.
  • If you save an execution time that has already passed for the current day, the rule first runs at that time on the next day.
  • You cannot use 0 as Target vCPU.
  • A rule runs only while the cluster is Running. If the cluster is not Running, VeloDB Cloud retries the scheduled execution for up to 30 minutes and then skips that execution.

Before you begin​

  • You need permission to manage compute resources in the warehouse.
  • The cluster must be Running.
  • Disable AutoScaling before you configure a time-based scaling policy.

Create and enable a policy​

  1. Log in to the VeloDB Cloud console.
  2. In the upper-left corner, select the warehouse that contains the cluster.
  3. In the left navigation pane, under Compute, click Clusters.
  4. Click Physical Clusters.
  5. Click Card, and then click a cluster in the Running state to open its details panel.
  6. Click Scaling.
  7. In Elastic Scaling, under Scaling Strategy, select Time-based Scaling.
  8. Click Add.
  9. Set Execution Time and Target vCPU for the rule.
  10. Click Save.
  11. Repeat steps 8 through 10 to add another rule. Set an Execution Time that is at least one hour apart from the first rule and a different Target vCPU.
  12. Turn on Time-based Scaling Policy.
  13. Review the scaling-in impact, and then click Confirm.

To verify the configuration, confirm that the policy toggle is on and that all saved rules appear in the rule table. In the cluster details panel, confirm that Scaling Mode is Time-based Scaling.

When a rule changes the cluster capacity, the cluster status changes to Scaling and returns to Running when the resource change finishes. Check the compute and cache values in the details panel to confirm that the target configuration took effect.

Edit a rule​

  1. Open Time-based Scaling for the cluster.
  2. Click Edit for the rule.
  3. Change Execution Time or Target vCPU.
  4. Click Save.

Confirm that the updated values appear in the rule table.

Disable a policy​

  1. Open Time-based Scaling for the cluster.
  2. Turn off Time-based Scaling Policy.
  3. In Disable time-based scaling policy?, click Confirm.

Saved rules remain in the table, but they do not run while the policy is disabled. In the cluster details panel, Scaling Mode changes to Manual Scaling.

Delete a rule​

  1. Click the delete icon for the rule.
  2. In Delete rule?, confirm the deletion.

The deleted rule no longer appears in the table and does not run at its former execution time. If fewer than two saved rules remain, the console automatically disables the policy. Add enough valid rules before you enable it again.

See also​