Skip to content

Latest commit

 

History

History
59 lines (49 loc) · 2.43 KB

elastic-agent-configuration.asciidoc

File metadata and controls

59 lines (49 loc) · 2.43 KB

Configure standalone {agent}s

Tip
 To get started quickly, use {kib} to create and download a standalone policy file. You’ll still need to deploy and manage the file, though. For more information, refer to [create-standalone-agent-policy] or try out our example: Use standalone {agent} to monitor nginx.

Standalone {agent}s are manually configured and managed locally on the systems where they are installed. They are useful when you are not interested in centrally managing agents in {fleet}, either due to your company’s security requirements, or because you prefer to use another configuration management system.

To configure standalone {agent}s, specify settings in the elastic-agent.yml policy file deployed with the agent. Prior to installation, the file is located in the extracted {agent} package. After installation, the file is copied to the directory described in [installation-layout]. To apply changes after installation, you must modify the installed file.

For installation details, refer to [install-standalone-elastic-agent].

Alternatively, you can put input configurations in YAML files into the folder {path.config}/inputs.d to separate your configuration into multiple smaller files. The YAML files in the inputs.d folder should contain input configurations only. Any other configurations are ignored. The files are reloaded at the same time as the standalone configuration.

Tip
 The first line of the configuration must be inputs. Then you can list the inputs you would like to run. Each input in the policy must have a unique value for the id key. If the id key is missing its value defaults to the empty string "". 
inputs:
  - id: unique-logfile-id
    type: logfile
    data_stream.namespace: default
    paths: [/path/to/file]
    use_output: default

  - id: unique-system-metrics-id
    type: system/metrics
    data_stream.namespace: default
    use_output: default
    streams:
      - metricset: cpu
        data_stream.dataset: system.cpu

The following sections describe some settings you might need to configure to run an {agent} standalone. For a full reference example, refer to the elastic-agent.reference.yml file.

The settings described here are available for standalone {agent}s. Settings for {fleet}-managed agents are specified through the UI. You do not set them explicitly in a policy file.