Major GitHub issue:
https://github.com/openshift/openshift-docs/issues/39533
Which section(s) is the issue in?
Json log formatter configuration examples are poor or missing whole information.
What needs fixing?
It is best to have full examples of full object yamls instead of the information cut in pieces.
Today it has this in the docs:
pipelines: - inputRefs: [ application ] outputRefs: myFluentd parse: json
or this:
outputDefaults: - elasticsearch: structuredTypeKey: kubernetes.labels.logFormat structuredTypeName: nologformat pipelines: - inputRefs: <application> outputRefs: default parse: json
But it should be something like this. (This was is better so we avoid users that will need to infer or guess where they should go or what object or what part in the object they will need to edit):
apiVersion: "logging.openshift.io/v1" kind: "ClusterLogForwarder" metadata: name: "instance" namespace: "openshift-logging" spec: pipeline: - inputRefs: [ application ] outputRefs: [ elastic ] outputs: - name: elastic type: elasticsearch url: http://elasticsearch.openshift-logging.svc.cluster.local:9200 elasticsearch: structuredTypeKey: kubernetes.label.appType