Scaling
Scaling changes the resources available to a compute cluster. VeloDB Cloud supports three scaling methods: manual scaling, time-based scaling, and auto-scaling. Choose one according to how predictable the workload is and how much control you want.
| Scaling method | Best for | How it works | What you configure |
|---|---|---|---|
| Manual scaling | Immediate, one-time resource changes | You manually choose the target compute or cache size. | Target compute size, or separate cache scale-out size. |
| Time-based scaling | Predictable and recurring business peaks and lows | The cluster scales at fixed daily times according to rules. | At least two scheduled rules with different target vCPU values. |
| Auto-scaling | Dynamic or hard-to-predict workload changes | VeloDB Cloud evaluates recent resource usage and automatically chooses a suitable CPU tier. | Auto-scaling switch, Min CPU Cores, and Max CPU Cores. |
Use manual scaling when you need an immediate, explicit resource change. For example, scale out compute before a known one-time workload, scale in compute after traffic drops, or scale out cache when the working set grows.
Use time-based scaling when workload peaks are predictable and repeat at known times. For example, configure daily rules for business hours and off-peak hours.
Use auto-scaling when workload changes are dynamic or hard to predict. Auto-scaling reacts to actual CPU and memory pressure within the min/max CPU cores that you configure.
Frontend Groups
Frontend Groups lets you scale the FE (Frontend) resource group independently of compute clusters. FE nodes handle query parsing, planning, and metadata management. This feature is currently in Private Preview.