ILM Policy variable conflicts with ILM Policies

[valencetech]

Describe the bug
When ILM is enabled, the fluentd pod goes into a crash loop because both ILM_POLICY and ILM_POLICIES are enabled even though the ILM_POLICIES is an empty map. Only one of the variables should actually be defined.

Version of Helm and Kubernetes:

Helm Version: 3.5.2

$ helm version
please put the output of it here

Kubernetes Version: 1.18.7

$ kubectl version
please put the output of it here

Which version of the chart: 11.13.0

What happened: Deployment fails due to Pod in crash loop. Two conflicting variables are defined instead of one.

What you expected to happen: I should be able to enable ILM and chose either to provide a policy definition or an existing policy name

How to reproduce it (as minimally and precisely as possible): Set ilm.enabled to true

<~--
This could be something like:

values.yaml (only put values which differ from the defaults)

key: value

helm install my-release {{ .GitHubOrg }}/name-of-chart --version version --values values.yaml

-->

Anything else we need to know:

[monotek]

Would you mind to create an PR with a fix?

[Truper4]

I am also having troubles with ILM enabled policy fluentd-elastisearch tag: 3.2.0 image and version: 11.7.2 of helm chart.
Configuration like this doesnt work:

logstash:
    enabled: true
    prefix: "logstash"
    prefixSeparator: "-"
    dateformat: "%Y.%m.%d"
  ilm:
    enabled: false
    policy_id: logstash-policy-by-fluentd
    policy:
      phases:
        hot:
          min_age: 0ms
          actions:
            rollover:
              max_age: 30d
              max_size: 100gb
            set_priority:
                priority: 100
        delete:
          min_age: 60d
          actions:
            delete:
    policies: {}
      # example for ilm policies config
      # ilm_policy_id1: {}
      # ilm_policy_id2: {}
    policy_overwrite: false
  template:
    enabled: false
    overwrite: false
    name: fluentd-template
    file: fluentd-template.json
    content: |-
      {
        "index_patterns": [
            "logstash-*"
        ],
        "settings": {
            "index": {
                "number_of_replicas": "1"
            }
        }
      }

  path: "/"

Error what I am getting after enabling ILM:
[error]: config error file="/etc/fluent/fluent.conf" error_class=Fluent::ConfigError error="xpack endpoint does not work"

I have checked if the xpack is installed and after i search for gem of xpack i can see it in the fluentd-elasticsearch container.

Also when I check if elasticsearch has working xpack all seems good.

[CoverID]

We can enable ILM with template like this

template:
    enabled: true
    overwrite: true
    useLegacy: true
    name: fluentd-template
    file: fluentd-template.json
    content: |-
      {
        "index_patterns": [
            "logstash-*"
        ],
        "settings": {
            "index": {
                "number_of_replicas": "1",
                "lifecycle": {
                  "name": "7-days-default",
                  "rollover_alias": "logstash-history"
                }
            }
        }
      }

*Note: you need to edit your Policy and disable Rollover at Hot phase, why? by default fluentd index pattern start with logstash-%Y.%m.%d and rollover_alias cannot use pattern like that