Resource Requirements | Rasa Documentation

Below are the recommended CPU and Memory requirements for the individual pods. You can experiment with different values according to your cluster usage metrics and adjust accordingly. These values can be set under the resources section under each component in the values.yaml file.

resources:
  limits:
    cpu: 2000m
    memory: 2000Mi
  requests:
    cpu: 1000m
    memory: 1000Mi

Pod	CPU	Memory
backend	1000m	2000Mi
keycloak	1000m	2000Mi
web-client	minimal/default	minimal/default
event-ingestion	500m	1000Mi
Rasa model-service	2000m	4000Mi

The resource requirements for the rasa model-service pod will depend on the number of trainings and active models that are being run. Training of a model normally requires around 800 MB of RAM and 1000m of CPU units. Allocate higher CPU resources if you anticipate a higher number of parallel trainings.

The resource requirements for the event-ingestion will depend on the load the production assistant expects. We recommend you to start with the default values and monitor the resource usage in your cluster.