Scaling

Zeabur allows you to adjust the resources allocated to your services so they can handle varying workloads. You can scale vertically by increasing CPU and memory, and choose the right infrastructure tier for your needs.

Vertical Scaling

Vertical scaling means increasing (or decreasing) the CPU and memory allocated to a single service instance.

To adjust resources:

  1. Open your service in the Zeabur dashboard.
  2. Go to Settings.
  3. Under the Resources section, configure the desired CPU and memory limits.
  4. Save your changes. The new resource allocation takes effect on the next deployment.

Resource Plans

Zeabur offers several resource tiers. The default allocation is suitable for most lightweight services, but you can increase resources for more demanding workloads such as databases, build servers, or AI inference.

💡

Resource allocation directly affects billing. Higher CPU and memory limits result in higher costs. See Pricing for details on per-unit pricing.

Dedicated Servers

For workloads that require guaranteed resources, predictable performance, or compliance with data residency requirements, Zeabur offers Dedicated Servers. A dedicated server gives you an isolated compute environment that is not shared with other users.

Benefits of dedicated servers include:

  • Guaranteed resources — No noisy-neighbor issues.
  • Custom regions — Deploy in specific geographic locations.
  • Higher resource ceilings — Run services that exceed shared cluster limits.

See Dedicated Servers for setup instructions and available configurations.

Scaling Strategies

ScenarioRecommendation
Occasional traffic spikesIncrease CPU/memory to handle peak load
Consistent high trafficUse a dedicated server for predictable performance
Cost optimizationRight-size resources based on metrics
Temporary inactivitySuspend the service to stop billing

Monitoring Resource Usage

Use the Metrics dashboard to monitor CPU, memory, and network usage in real time. This data helps you determine when to scale up or down. For cost analysis, check the Service Usage page.