This documentation contains work-in-progress information for future Elastic Stack and Cloud releases. Use the version selector to view supported release docs. It also contains some Elastic Cloud serverless information. Check out our serverless docs for more details.

« Config file example: Nginx HTTP Server Example: Use standalone Elastic Agent with Elastic Cloud Serverless to monitor nginx »

› ›

Grant standalone Elastic Agents access to Elasticsearch

edit

Grant standalone Elastic Agents access to Elasticsearch

edit

You can use either API keys or user credentials to grant standalone Elastic Agents access to Elasticsearch resources. The following minimal permissions are required to send logs, metrics, traces, and synthetics to Elasticsearch:

monitor cluster privilege
auto_configure and create_doc index privileges on logs-*-*, metrics-*-*, traces-*-*, and synthetics-*-*.

It’s recommended that you use API keys to avoid exposing usernames and passwords in configuration files.

If you’re using Fleet, refer to Fleet enrollment tokens.

Create API keys for standalone agents

edit

API keys are sent as plain-text, so they only provide security when used in combination with Transport Layer Security (TLS). Our hosted Elasticsearch Service on Elastic Cloud provides secure, encrypted connections out of the box! For self-managed Elasticsearch clusters, refer to Public Key Infrastructure (PKI) certificates.

You can set API keys to expire at a certain time, and you can explicitly invalidate them. Any user with the manage_api_key or manage_own_api_key cluster privilege can create API keys.

For security reasons, we recommend using a unique API key per Elastic Agent. You can create as many API keys per user as necessary.

If you are using Elastic Cloud Serverless, API key authentication is required.

To create an API key for Elastic Agent:

In an Elastic Cloud or on premises environment, in Kibana navigate to Stack Management > API keys and click Create API key.

In a Serverless environment, in Kibana navigate to Project settings > Management > API keys and click Create API key.
Enter a name for your API key and select Control security privileges.
In the role descriptors box, copy and paste the following JSON. This example creates an API key with privileges for ingesting logs, metrics, traces, and synthetics:
```
{
  "standalone_agent": {
    "cluster": [
      "monitor"
    ],
    "indices": [
      {
        "names": [
          "logs-*-*", "metrics-*-*", "traces-*-*", "synthetics-*-*" 
        ],
        "privileges": [
          "auto_configure", "create_doc"
        ]
      }
    ]
  }
}
```
Adjust this list to match the data you want to collect. For example, if you aren’t using APM or synthetics, remove "traces-*-*" and "synthetics-*-*" from this list.
To set an expiration date for the API key, select Expire after time and input the lifetime of the API key in days.
Click Create API key.

You’ll see a message indicating that the key was created, along with the encoded key. By default, the API key is Base64 encoded, but that won’t work for Elastic Agent.

Click the down arrow next to Base64 and select Beats.
Copy the API key. You will need this for the next step, and you will not be able to view it again.

To use the API key, specify the api_key setting in the elastic-agent.yml file. For example:

[...]
outputs:
  default:
    type: elasticsearch
    hosts:
      - 'https://da4e3a6298c14a6683e6064ebfve9ace.us-central1.gcp.cloud.es.io:443'
    api_key: _Nj4oH0aWZVGqM7MGop8:349p_U1ERHyIc4Nm8_AYkw 
[...]

The format of this key is <id>:<key>. Base64 encoded API keys are not currently supported in this configuration.

For more information about creating API keys in Kibana, see API Keys.

Create a standalone agent role

edit

Although it’s recommended that you use an API key instead of a username and password to access Elasticsearch (and an API key is required in a Serverless environment), you can create a role with the required privileges, assign it to a user, and specify the user’s credentials in the elastic-agent.yml file.

In Kibana, go to Stack Management > Roles.
Click Create role and enter a name for the role.
In Cluster privileges, enter monitor.
In Index privileges, enter:
1. logs-*-*, metrics-*-*, traces-*-* and synthetics-*-* in the Indices field.
  
  Adjust this list to match the data you want to collect. For example, if you aren’t using APM or synthetics, remove traces-*-* and synthetics-*-* from this list.
2. auto_configure and create_doc in the Privileges field.
Create the role and assign it to a user. For more information about creating roles, refer to Kibana role management.
To use these credentials, set the username and password in the elastic-agent.yml file:
```
[...]
outputs:
  default:
    type: elasticsearch
    hosts:
      - 'https://da4e3a6298c14a6683e6064ebfve9ace.us-central1.gcp.cloud.es.io:443'
    username: ES_USERNAME 
    password: ES_PASSWORD
[...]
```
For security reasons, specify a user with the minimal privileges described here. It’s recommended that you do not use the elastic superuser.

« Config file example: Nginx HTTP Server Example: Use standalone Elastic Agent with Elastic Cloud Serverless to monitor nginx »