detectExceptions is not forbidden in multi-worker setups

[florianstoeber]

Describe the bug:
If I have a logging-instance running with multiple fluentd workers it is not forbidden to use the detectExceptions filter, while this will break the fluentd processes.

Expected behaviour:
As this is a known limitation, it would be nice to mark the Flow as "not valid".

Steps to reproduce the bug:
Create a logging-instance with multiple workers and create a Flow with detectExceptions filter.

Environment details:

• Kubernetes version (e.g. v1.15.2): 1.25
• Cloud-provider/provisioner (e.g. AKS, GKE, EKS, PKE etc): AKS
• logging-operator version (e.g. 2.1.1): 4.2.2
• Install method (e.g. helm or static manifests): Helm
• Logs from the misbehaving component (and any other relevant logs):
• Resource definition (possibly in YAML format) that caused the issue, without sensitive data:

apiVersion: logging.banzaicloud.io/v1beta1
kind: Flow
metadata:
  name: flow
spec:
  filters:
  - record_modifier:
      records:
      - fluentd_worker: ${ENV['HOSTNAME']}
  - record_transformer:
      remove_keys: $.kubernetes.docker_id, $.kubernetes.annotations, $.kubernetes.container_hash,
        $.kubernetes.pod_id
  - parser:
      key_name: message
      parse:
        type: json
      remove_key_name_field: true
      reserve_data: true
  - detectExceptions:
      languages:
      - java
      - python
      multiline_flush_interval: "0.1"
  localOutputRefs:
  - output
  match:
  - select: {}

/kind bug

[pepov]

The request is valid, although this is not a bug, but an enhancement request. It adds a validation on top of fluentd that doesn't tell us about the problem exactly.