IMPORTANT: No additional bug fixes or documentation updates will be released for this version. For the latest information, see the current release documentation.
« Kubernetes autodiscovery with Elastic Agent Hints annotations based autodiscover »
Elastic Docs ›Fleet and Elastic Agent Guide [8.6] ›Configure standalone Elastic Agents ›Kubernetes autodiscovery with Elastic Agent

Conditions based autodiscover

edit

Conditions based autodiscover

edit

You can define autodiscover conditions in each input to allow Elastic Agent to automatically identify Pods and start monitoring them using predefined integrations. Refer to Configure inputs for Standalone Elastic Agents to get an idea .

Example: Target Pods by label

edit

To automatically identify a Redis Pod and monitor it with the Redis integration, uncomment the following input configuration inside the Elastic Agent Standalone manifest:

- name: redis
  type: redis/metrics
  use_output: default
  meta:
    package:
      name: redis
      version: 0.3.6
  data_stream:
    namespace: default
  streams:
    - data_stream:
        dataset: redis.info
        type: metrics
      metricsets:
        - info
      hosts:
        - '${kubernetes.pod.ip}:6379'
      idle_timeout: 20s
      maxconn: 10
      network: tcp
      period: 10s
      condition: ${kubernetes.labels.app} == 'redis'

The condition ${kubernetes.labels.app} == 'redis' will make the Elastic Agent look for a Pod with the label app:redis within the scope defined in its manifest.

For a list of provider fields that you can use in conditions, refer to Kubernetes Provider.

The redis input defined in the Elastic Agent manifest only specifies the`info` metricset. To learn about other available metricsets and their configuration settings, refer to the Redis module page.

To deploy Redis, you can apply the following example manifest:

apiVersion: v1
kind: Pod
metadata:
  name: redis
  labels:
    k8s-app: redis
    app: redis
spec:
  containers:
  - image: redis
    imagePullPolicy: IfNotPresent
    name: redis
    ports:
    - name: redis
      containerPort: 6379
      protocol: TCP

You should now be able to see Redis data flowing in on index metrics-redis.info-default. Make sure the port in your Redis manifest file matches the port used in the Redis input.

All assets (dashboards, ingest pipelines, and so on) related to the Redis integration are not installed. You need to explicitly install them through Kibana.

To set the target host dynamically for a targeted Pod based on its labels, use a variable in the Elastic Agent policy to return path information from the provider:

- data_stream:
    dataset: kubernetes.scheduler
    type: metrics
  metricsets:
    - scheduler
  bearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/token
  hosts:
    - 'https://${kubernetes.pod.ip}:10259'
  period: 10s
  ssl.verification_mode: none
  condition: ${kubernetes.labels.component} == 'kube-scheduler'

In some "As a Service" Kubernetes implementations, like GKE, the control plane nodes or even the Pods running on them won’t be visible. In these cases, it won’t be possible to use scheduler metricsets, necessary for this example. Refer scheduler and controller manager to find more information.

Following the Redis example, if you deploy another Redis Pod with a different port, it should be detected. To check this, go, for example, to the field service.address under metrics-redis.info-default. It should be displaying two different services.

To obtain the policy generated by this configuration, connect to Elastic Agent container:

kubectl exec -n kube-system --stdin --tty elastic-agent-standalone-id -- /bin/bash

Do not forget to change the elastic-agent-standalone-id to your Elastic Agent Pod’s name. Moreover, make sure that your Pod is inside kube-system. If not, change -n kube-system to the correct namespace.

Inside the container inspect the output of the configuration file you used for the Elastic Agent:

elastic-agent inspect --variables --variables-wait 1s -c /etc/elastic-agent/agent.yml
You should now be able to see the generated policy. If you look for the scheduler, it will look similar to this. 
- bearer_token_file: /var/run/secrets/kubernetes.io/serviceaccount/token
  hosts:
    - https://172.19.0.2:10259
  index: metrics-kubernetes.scheduler-default
  meta:
    package:
      name: kubernetes
      version: 1.9.0
  metricsets:
    - scheduler
  module: kubernetes
  name: kubernetes-node-metrics
  period: 10s
  processors:
    - add_fields:
        fields:
          labels:
            component: kube-scheduler
            tier: control-plane
          namespace: kube-system
          namespace_labels:
            kubernetes_io/metadata_name: kube-system
          namespace_uid: 03d6fd2f-7279-4db4-9a98-51e50bbe5c62
          node:
            hostname: kind-control-plane
            labels:
              beta_kubernetes_io/arch: amd64
              beta_kubernetes_io/os: linux
              kubernetes_io/arch: amd64
              kubernetes_io/hostname: kind-control-plane
              kubernetes_io/os: linux
              node-role_kubernetes_io/control-plane: ""
              node_kubernetes_io/exclude-from-external-load-balancers: ""
            name: kind-control-plane
            uid: b8d65d6b-61ed-49ef-9770-3b4f40a15a8a
          pod:
            ip: 172.19.0.2
            name: kube-scheduler-kind-control-plane
            uid: f028ad77-c82a-4f29-ba7e-2504d9b0beef
        target: kubernetes
    - add_fields:
        fields:
          cluster:
            name: kind
            url: kind-control-plane:6443
        target: orchestrator
    - add_fields:
        fields:
          dataset: kubernetes.scheduler
          namespace: default
          type: metrics
        target: data_stream
    - add_fields:
        fields:
          dataset: kubernetes.scheduler
          target: event
    - add_fields:
        fields:
          id: ""
          snapshot: false
          version: 8.3.0
        target: elastic_agent
    - add_fields:
        fields:
          id: ""
        target: agent
  ssl.verification_mode: none

Example: Dynamic logs path

edit

To set the log path of Pods dynamically in the configuration, use a variable in the Elastic Agent policy to return path information from the provider:

- name: container-log
  id: container-log-${kubernetes.pod.name}-${kubernetes.container.id}
  type: filestream
  use_output: default
  meta:
    package:
      name: kubernetes
      version: 1.9.0
  data_stream:
    namespace: default
  streams:
    - data_stream:
      dataset: kubernetes.container_logs
      type: logs
      prospector.scanner.symlinks: true
      parsers:
        - container: ~
      paths:
        - /var/log/containers/*${kubernetes.container.id}.log
The policy generated by this configuration will look similar to this for every Pod inside the scope defined in the manifest. 
- id: container-log-etcd-kind-control-plane-af311067a62fa5e4d6e5cb4d31e64c1c35d82fe399eb9429cd948d5495496819
  index: logs-kubernetes.container_logs-default
  meta:
    package:
      name: kubernetes
      version: 1.9.0
  name: container-log
  parsers:
    - container: null
  paths:
    - /var/log/containers/*af311067a62fa5e4d6e5cb4d31e64c1c35d82fe399eb9429cd948d5495496819.log
  processors:
    - add_fields:
        fields:
          id: af311067a62fa5e4d6e5cb4d31e64c1c35d82fe399eb9429cd948d5495496819
          image:
            name: registry.k8s.io/etcd:3.5.4-0
          runtime: containerd
        target: container
    - add_fields:
        fields:
          container:
            name: etcd
        labels:
          component: etcd
          tier: control-plane
        namespace: kube-system
        namespace_labels:
          kubernetes_io/metadata_name: kube-system
        namespace_uid: 03d6fd2f-7279-4db4-9a98-51e50bbe5c62
        node:
          hostname: kind-control-plane
          labels:
            beta_kubernetes_io/arch: amd64
            beta_kubernetes_io/os: linux
            kubernetes_io/arch: amd64
            kubernetes_io/hostname: kind-control-plane
            kubernetes_io/os: linux
            node-role_kubernetes_io/control-plane: ""
            node_kubernetes_io/exclude-from-external-load-balancers: ""
          name: kind-control-plane
          uid: b8d65d6b-61ed-49ef-9770-3b4f40a15a8a
        pod:
          ip: 172.19.0.2
          name: etcd-kind-control-plane
          uid: 08970fcf-bb93-487e-b856-02399d81fb29
      target: kubernetes
    - add_fields:
        fields:
          cluster:
            name: kind
            url: kind-control-plane:6443
        target: orchestrator
    - add_fields:
        fields:
          dataset: kubernetes.container_logs
          namespace: default
          type: logs
        target: data_stream
    - add_fields:
        fields:
          dataset: kubernetes.container_logs
        target: event
    - add_fields:
        fields:
          id: ""
          snapshot: false
          version: 8.3.0
        target: elastic_agent
    - add_fields:
        fields:
          id: ""
        target: agent
  prospector.scanner.symlinks: true
  type: filestream
« Kubernetes autodiscovery with Elastic Agent Hints annotations based autodiscover »