The resource requirements and limits of your Logging operator deployment must match the size of your cluster and the logging workloads. By default, the Logging operator uses the following configuration.

Fluent Bit 🔗︎

 - Limits:
   - cpu: 200m
   - memory: 100M
 - Requests:
   - cpu: 100m
   - memory: 50M

Fluentd 🔗︎

 - Limits:
   - cpu: 1
   - memory: 200M
 - Requests:
   - cpu: 500m
   - memory:  100M

You can easily change this in the Logging custom resource, for example:

apiVersion: logging.banzaicloud.io/v1beta1
kind: Logging
metadata:
  name: default-logging
  namespace: logging
spec:
  fluentd:
    resources:
      requests:
        cpu: 1
        memory: 1Gi
      limits:
        cpu: 2
        memory: 2Gi
  fluentbit:
    resources:
      requests:
        cpu: 500m
        memory: 500M
      limits:
        cpu: 1
        memory: 1Gi