Manage Clusters
A physical cluster is a compute group inside a VeloDB Cloud warehouse. Clusters in the same warehouse share stored data, while each cluster has independent compute resources, cache capacity, and cached data. Use multiple clusters to isolate workloads with different resource and performance requirements.
Open the Clusters page
- Log in to the VeloDB Cloud console.
- In the upper-left corner, select the warehouse to manage.
- In the left navigation pane, under Compute, click Clusters.
Cluster types
The Clusters page displays the following cluster types:
| Cluster type | Description |
|---|---|
| Physical cluster | Provides compute resources for workloads and local cache capacity. Use a physical cluster directly, or mount it as the active or standby cluster of a virtual cluster. |
| Virtual cluster | Pairs an active physical cluster and a standby physical cluster in different availability zones. A virtual cluster has no compute or cache resources of its own. |
The Virtual Clusters tab appears only when the selected warehouse supports virtual clusters. For requirements and failover behavior, see Create and manage a virtual cluster.
Choose a cluster view
Click Physical Clusters. Then select one of the following views:
| View | Use it to |
|---|---|
| Card | Select one cluster at a time and view its resource metrics, configuration, billing method, and lifecycle actions in the details panel. |
| Table | Compare multiple clusters, search by cluster name or ID, filter by status, and customize the displayed columns. |
Manage the cluster lifecycle
- Create a cluster to add compute capacity for a workload.
- View and manage clusters to compare clusters, review resource metrics and billing methods, or rename a cluster.
- Pause and resume a cluster to stop idle compute and resume it when needed.
- Restart a cluster to restart its compute resources.
- Delete a cluster when its compute and cache resources are no longer required.
Manage cluster capacity and workloads
- Scale a cluster manually, on a schedule, or automatically as workload demand changes.
- Configure Workload Groups to control how queries share resources within a cluster.
- Manage warmup jobs to preload data into a cluster's cache before workloads need it.