fluent-plugin-tkgi-metadata-parser

Description

This Fluentd parser plugin parses metadata received from Tanzu Kubernetes Grid Integrated Edition (TKGI) or Tanzu Kubernetes Grid (TKG).

TKG(I) uses RFC 5424 - The Syslog Protocol, this plugin parses specifically the syslog5424_sd field if you use standard parser to parse syslog5424 logs initially, but you can use any field which consists of Kubernetes metadata.

e.g

"syslog5424_sd":"[kubernetes@47450 pod-template-hash=\"6cdc894687\" app.kubernetes.io/component=\"operator\" app.kubernetes.io/instance=\"prometheus\" app.kubernetes.io/managed-by=\"Helm\" app.kubernetes.io/name=\"kube-prometheus\" helm.sh/chart=\"kube-prometheus-3.2.0\" namespace_name=\"prometheus\" object_name=\"prometheus-kube-prometheus-operator-6cdc894687-c56bf\" container_name=\"prometheus-operator\" vm_id=\"04c2872a-768a-4577-944a-87842407f582\"]"

Installation

$ td-agent-gem install fluent-plugin-tkgi-metadata-parser

Usage

To parse log initially when it comes to Fluentd you need to apply syslog5424 parser, example below uses grok parser plugin but feel free to use any other parser of your choice.

<source>
    @type tcp
    port 6514
    key_name message
    <parse>
        @type grok
        grok_failue_key _grokparsefailure
        <grok>
            pattern %{SYSLOG5424PRI}%{NONNEGINT:syslog5424_ver} +(?:%{TIMESTAMP_ISO8601:syslog5424_ts}|-) +(?:%{HOSTNAME:syslog5424_host}|-) +(?:%{NOTSPACE:syslog5424_app}|-) +(?:%{NOTSPACE:syslog5424_proc}|-) +(?:%{WORD:syslog5424_msgid}|-) +(?:%{SYSLOG5424SD:syslog5424_sd}|-|) +%{GREEDYDATA:syslog5424_msg}
        </grok>
    </parse>
    tag tkgi
</source>

Plugin configuration

<filter tkgi>
    @type parser
    key_name syslog5424_sd
    reserve_data true
    reserve_time true
    <parse>
      @type tkgi_metadata
      es_mode true
    </parse>
</filter>

Once parsed the output looks like this:

{
  "_index": "k8s-2020.12.31",
  "_type": "_doc",
  "_id": "2zqJunYBHg9f6MMRmev2",
  "_version": 1,
  "_score": null,
  "_source": {
    "syslog5424_pri": "14",
    "syslog5424_ver": "1",
    "syslog5424_ts": "2020-12-31T20:41:24.224539+00:00",
    "syslog5424_host": "XXXX",
    "syslog5424_app": "pod.log/prometheus/prometheus-kube-prometheus-op",
    "syslog5424_proc": "-",
    "syslog5424_msg": "something very important",
    "app_kubernetes_io/component": "operator",
    "app_kubernetes_io/instance": "prometheus",
    "app_kubernetes_io/managed-by": "Helm",
    "app_kubernetes_io/name": "kube-prometheus",
    "helm_sh/chart": "kube-prometheus-3.2.0",
    "pod-template-hash": "6cdc894687",
    "namespace_name": "prometheus",
    "object_name": "prometheus-kube-prometheus-operator-6cdc894687-c56bf",
    "container_name": "prometheus-operator",
    "vm_id": "04c2872a-768a-4577-944a-87842407f582",
    "source": "kubernetes",
    "source_id": "47450",
    "@timestamp": "2020-12-31T14:41:25.003082896-06:00",
    "tag": "tkgi"
  }

Optional Parameters

• delimiter

  delimiter which separate each key-value pairs. whitespaces or tabs can be given in quotes: ie, "" or "\t" . By default it is "".

• es_mode

  When using Elasticsearch as a storage database, it expects mapping to be an object and not text, due to Kubernetes label/annotation naming conventions you might get the following error:

  <Fluent::Plugin::ElasticsearchErrorHandler::ElasticsearchError: 400 - Rejected by Elasticsearch [error type]: mapper_parsing_exception [reason]: 'Could not dynamically add mapping for field [app.kubernetes.io/component]. Existing mapping for [app] must be of type object but found [text].'> 
  

  This issue is caused by label/annotation dots (.) creating hierarchy in Elasticsearch documents. If an annotation has a structure like this: example.annotation.data: some-data and a different one contains example.annotation: value their mapping will conflict, as example.annotation is both an object and a keyword.

  To avoid this, when set to true this plugin replaces (.) with (_)

  By default, it is false

Contributing

1. Fork it
2. Create your feature branch (git checkout -b my-new-feature)
3. Commit your changes (git commit -am 'Add some feature')
4. Push to the branch (git push origin my-new-feature)
5. Create a new Pull Request

TODO:

Remove grok parser dependency to parse syslog initially.