contour_cpu_request
Description
The contour_cpu_request parameter sets the CPU request for the Contour control-plane Pods on a Rack. It applies only when router_type=contour. The value is a standard Kubernetes CPU quantity (for example 100m or 250m).
This parameter has no effect on nginx Racks. On a Rack with router_type=nginx, the value is ignored regardless of what it is set to.
Default Value
The default value is 100m. This default preserves existing behavior: a Rack that has not set contour_cpu_request schedules the Contour control plane with a 100m CPU request, the same value used before this parameter was available.
Use Cases
- Raise the request when the Contour control plane is CPU constrained, for example on a Rack with many routes where reconciliation is slow.
- Lower the request on small or low-traffic Racks to free CPU for other workloads on the same nodes.
Setting the Parameter
$ convox rack params set contour_cpu_request=250m -r rackName
Setting parameters... OK
Viewing Current Configuration
$ convox rack params -r rackName
Additional Information
This parameter is available on AWS Racks only and takes effect when router_type=contour.
contour_cpu_request is clearable. Clearing it (setting it to an empty value) restores the default of 100m. Setting it to a new value reschedules the Contour control-plane Pods with the updated CPU request.
Because changing this value reschedules the control-plane Pods, apply it during a window where a brief Contour control-plane restart is acceptable.
See Also
