Fluentd Plugin to re-tag based on log metadata 🔗︎

Overview 🔗︎

More info at https://github.com/banzaicloud/fluent-plugin-tag-normaliser

Available kubernetes metadata

Parameter Description Example
${pod_name} Pod name understood-butterfly-logging-demo-7dcdcfdcd7-h7p9n
${container_name} Container name inside the Pod logging-demo
${namespace_name} Namespace name default
${pod_id} Kubernetes UUID for Pod 1f50d309-45a6-11e9-b795-025000000001
${labels} Kubernetes Pod labels. This is a nested map. You can access nested attributes via . {“app”:“logging-demo”, “pod-template-hash”:“7dcdcfdcd7” }
${host} Node hostname the Pod runs on docker-desktop
${docker_id} Docker UUID of the container 3a38148aa37aa3…

Configuration 🔗︎

Tag Normaliser parameters 🔗︎

Variable Name Type Required Default Description
format string No ${namespace_name}.${pod_name}.${container_name} Re-Tag log messages info at github

Example Parser filter configurations 🔗︎

apiVersion: logging.banzaicloud.io/v1beta1
kind: Flow
metadata:
 name: demo-flow
spec:
 filters:
   - tag_normaliser:
       format: cluster1.${namespace_name}.${pod_name}.${labels.app}
 selectors: {}
 outputRefs:
   - demo-output

Fluentd Config Result 🔗︎

<match kubernetes.**>
 @type tag_normaliser
 @id test_tag_normaliser
 format cluster1.${namespace_name}.${pod_name}.${labels.app}
</match>