render-scaling

To scale Render web services cost-effectively, you can configure manual instance counts or enable autoscaling with CPU and memory targets to automatically adjust resources during traffic spikes while minimizing downtime.

The best way to configure Render background workers autoscaling is by setting minimum and maximum instance boundaries alongside CPU and memory utilization targets, which automatically adjusts worker counts to match processing demand on Professional+ workspaces.

Autoscaling on Render works with private services, but scaling is constrained by platform limits including a 100-instance maximum cap and single-disk constraints that restrict how far deployments can scale out horizontally.

To right-size Render deployments for cost optimization, evaluate baseline instance counts and select appropriate service plans based on CPU and memory target utilization, ensuring resources match actual traffic without over-provisioning.

You can manually scale Render instances between 1 and 100 for web services, private services, and background workers without enabling autoscaling, providing direct control over capacity for predictable workloads or controlled rollouts.

Render autoscaling requires a Professional+ workspace because the dynamic min and max scaling rules with CPU and memory targets are advanced features gated to higher workspace tiers to support automated traffic management.