Plans of a service
The plan of a service gives the memory limit and the CPU limit of each instance and each job run.
services:
- type: web
name: hello
numInstances: 2
plan: starter # 512 MiB / 0.5 CPU per instanceA plan is an instance type of Render. These plans are for web services, private services, background workers and cron jobs.
The plans
plan | Memory | CPU |
|---|---|---|
free, starter | 512 MiB | 0.5 |
standard | 2 GiB | 1 |
pro | 4 GiB | 2 |
pro plus | 8 GiB | 4 |
pro max | 16 GiB | 4 |
pro ultra | 32 GiB | 8 |
The limits apply to each instance and to each job run.
How Ferry reads the name
Ferry ignores the case and the separator. pro plus, Pro_Plus, pro-plus and proplus are the same plan.
A plan that Ferry does not know
The plan sets nothing. The warning lists the plans for that type of entry:
service 'api': ignoring unknown plan 'enterprise' (known plans: free, starter, standard, pro, pro plus, pro max, pro ultra); set memoryLimit / cpuLimit insteadOn Render, the free plan has 0.1 CPU. Ferry gives it 0.5 CPU, the same as starter. On your own server, a tenth of a core only makes a service slow to start and to pass its health check. It makes nothing free for the host.
Static sites have no plans on Render. Thus Ferry ignores their plan, with no warning. memoryLimit and cpuLimit apply to them, because Ferry serves a static site from a container.