Skip to main content
This page is the single source of truth for limits applied at each subscription tier. Limits apply at three levels: per-organization caps on team and project counts, per-service caps on a single service, and organization-wide pools that bound total resource usage across every service in your org. For pricing, see suga.app/pricing.

​
Plan Summary

​
Per-Service Limits

Each service can be configured up to these tier-specific maximums.
Replicas multiply resource usage. A service configured with 2 CPU cores and 4 replicas counts as 8 CPU cores against your organization’s pool.

​
Organization Resource Pools

Beyond per-service caps, every organization has a shared resource budget across all services in all environments. Deployments are blocked when they would exceed any pool limit. For services with autoscaling enabled, the maximum values count toward pool usage. Suga reserves the burst ceiling rather than the baseline.

​
Volume Limits

Persistent volumes have separate per-volume and per-environment caps. Volume storage counts toward your organization’s overall storage pool.

​
CPU and Memory Ratio

Suga Cloud requires per-service memory to fall within 1 GiB to 6.5 GiB per CPU core. This applies to every service, on every tier.
If you request CPU and memory outside this ratio, Suga Cloud rounds up the smaller resource to bring it back in range, which may result in higher than expected billing. Always configure CPU and memory together.
For autoscaling services, the ratio applies to both the minimum and maximum bounds.

​
Vertical Autoscaling

Enable autoscaling on a service to let Suga Cloud adjust CPU and memory based on actual usage, between the limits you set.

​
How It Works

You configure two values per resource:
  • Minimum: the baseline allocation, always reserved for the service
  • Maximum: the ceiling Suga will scale to under load
Suga Cloud monitors usage and adjusts allocation within this range, scaling up as sustained usage approaches the current allocation and scaling down when usage drops well below it.

​
Resizing Behavior

When possible, Suga Cloud resizes resources in place without restarting your service. If the OS or runtime requires a restart to pick up the new memory limit, the service is restarted with the new allocation.
Services should still handle restarts: drain in-flight requests and persist state to volumes.

​
Pool Accounting

For autoscaling services, maximum values count toward your organization’s resource pool. This reserves scale-up headroom against the pool budget. A Pro org with a 100 CPU pool and four services configured at max 25 cores has used its full pool budget, even when those services are running at their minimum.

​
What Happens at a Limit

To increase your limits, upgrade your plan or contact us about Enterprise pricing.