Compute Plans & SLAs¶
Compute plans define the resource allocation and service level for your instances. They determine how much CPU and memory your service gets, and what level of availability guarantee you receive.
Compute plans¶
A compute plan is a resource tier that specifies:
- CPU -- the amount of processing power allocated
- Memory -- the amount of RAM allocated
When provisioning a service, you select a compute plan that matches your workload requirements. Larger plans provide more resources for demanding workloads.
Not all compute plans are available for every service offering. The available plans depend on the service, cloud provider, and zone.
SLA levels¶
Each compute plan assignment comes with a service level agreement (SLA) that defines the availability guarantee:
| SLA Level | Description |
|---|---|
| Best Effort | No formal availability guarantee. Suitable for development and testing workloads. |
| Guaranteed | Formal availability guarantee with defined uptime targets. Suitable for production workloads. |
The SLA level affects pricing -- guaranteed plans cost more than best-effort plans for the same resource tier.
Choosing a compute plan¶
Consider the following when selecting a compute plan:
- Workload type -- development/testing vs. production
- Resource needs -- how much CPU and memory does your application require?
- Availability requirements -- do you need a formal SLA?
- Budget -- higher tiers and guaranteed SLAs come at a higher cost
You can change the compute plan of an existing instance if your requirements change.