Use {{metricbeat}} to send monitoring data [monitoring-metricbeat-collection]

In 7.3 and later, you can use {{metricbeat}} to collect data about Metricbeat and ship it to the monitoring cluster. The benefit of using {{metricbeat}} instead of internal collection is that the monitoring agent remains active even if the Metricbeat instance dies.

Because you’ll be using {{metricbeat}} to monitor Metricbeat, you’ll need to run two instances of Metricbeat: a main instance that collects metrics from the system and services running on the server, and a second instance that collects metrics from Metricbeat only. Using a separate instance as a monitoring agent allows you to send monitoring data to a dedicated monitoring cluster. If the main agent goes down, the monitoring agent remains active.

If you’re running Metricbeat as a service, this approach requires extra work because you need to run two instances of the same installed service concurrently. If you don’t want to run two instances concurrently, use internal collection instead of using {{metricbeat}}.

To collect and ship monitoring data:

Configure the shipper you want to monitor
Install and configure {{metricbeat}} to collect monitoring data

Configure the shipper you want to monitor [configure-shipper]

Enable the HTTP endpoint to allow external collection of monitoring data:

Add the following setting in the Metricbeat configuration file (metricbeat.yml):
yaml
```
http.enabled: true
```
By default, metrics are exposed on port 5066. If you need to monitor multiple {{beats}} shippers running on the same server, set http.port to expose metrics for each shipper on a different port number:
yaml
```
http.port: 5067
```
Disable the default collection of Metricbeat monitoring metrics.

Add the following setting in the Metricbeat configuration file (metricbeat.yml):
yaml
```
monitoring.enabled: false
```
For more information, see Monitoring configuration options.
Configure host (optional).

If you intend to get metrics using {{metricbeat}} installed on another server, you need to bind the Metricbeat to host’s IP:
yaml
```
http.host: xxx.xxx.xxx.xxx
```
Configure cluster UUID.

The cluster UUID is necessary if you want to see {{beats}} monitoring in the {{kib}} stack monitoring view. The monitoring data will be grouped under the cluster for that UUID. To associate Metricbeat with the cluster UUID, set:
yaml
```
monitoring.cluster_uuid: "cluster-uuid"
```
Start Metricbeat.

Install and configure {{metricbeat}} to collect monitoring data [configure-metricbeat]

The next step depends on how you want to run {{metricbeat}}:
- If you’re running as a service and want to run a separate monitoring instance, take the the steps required for your environment to run two instances of {{metricbeat}} as a service. The steps for doing this vary by platform and are beyond the scope of this documentation.
- If you’re running the binary directly in the foreground and want to run a separate monitoring instance, install {{metricbeat}} to a different path. If necessary, set path.config, path.data, and path.log to point to the correct directories. See Directory layout for the default locations.
Enable the beat-xpack module in {{metricbeat}}.

For example, to enable the default configuration in the modules.d directory, run the following command, using the correct command syntax for your OS:
sh
```
metricbeat modules enable beat-xpack
```
For more information, see Configure modules and beat module.
Configure the beat-xpack module in {{metricbeat}}.

The modules.d/beat-xpack.yml file contains the following settings:
yaml
```
- module: beat
  metricsets:
    - stats
    - state
  period: 10s
  hosts: ["http://localhost:5066"]
  #username: "user"
  #password: "secret"
  xpack.enabled: true
```
Set the hosts, username, and password settings as required by your environment. For other module settings, it’s recommended that you accept the defaults.

By default, the module collects Metricbeat monitoring data from localhost:5066. If you exposed the metrics on a different host or port when you enabled the HTTP endpoint, update the hosts setting.

To monitor multiple {{beats}} agents, specify a list of hosts, for example:
yaml
```
hosts: ["http://localhost:5066","http://localhost:5067","http://localhost:5068"]
```
If you configured Metricbeat to use encrypted communications, you must access it via HTTPS. For example, use a hosts setting like https://localhost:5066.

If the Elastic {{security-features}} are enabled, you must also provide a user ID and password so that {{metricbeat}} can collect metrics successfully:
1. Create a user on the {{es}} cluster that has the remote_monitoring_collector built-in role. Alternatively, if it’s available in your environment, use the remote_monitoring_user built-in user.
2. Add the username and password settings to the beat module configuration file.
Optional: Disable the system module in the {{metricbeat}}.

By default, the system module is enabled. The information it collects, however, is not shown on the Stack Monitoring page in {{kib}}. Unless you want to use that information for other purposes, run the following command:
sh
```
metricbeat modules disable system
```
Identify where to send the monitoring data.

::::{tip} In production environments, we strongly recommend using a separate cluster (referred to as the monitoring cluster) to store the data. Using a separate monitoring cluster prevents production cluster outages from impacting your ability to access your monitoring data. It also prevents monitoring activities from impacting the performance of your production cluster. ::::

For example, specify the {{es}} output information in the {{metricbeat}} configuration file (metricbeat.yml):
yaml
```
output.elasticsearch:
  # Array of hosts to connect to.
  hosts: ["http://es-mon-1:9200", "http://es-mon2:9200"] <1>

  # Optional protocol and basic auth credentials.
  #protocol: "https"
  #api_key:  "id:api_key" <2>
  #username: "elastic"
  #password: "changeme"
```
1. In this example, the data is stored on a monitoring cluster with nodes es-mon-1 and es-mon-2.
2. Specify one of api_key or username/password.
If you configured the monitoring cluster to use encrypted communications, you must access it via HTTPS. For example, use a hosts setting like https://es-mon-1:9200.

::::{important} The {{es}} {{monitor-features}} use ingest pipelines. The cluster that stores the monitoring data must have at least one node with the ingest role. ::::

If the {{es}} {{security-features}} are enabled on the monitoring cluster, you must provide a valid user ID and password so that {{metricbeat}} can send metrics successfully:
1. Create a user on the monitoring cluster that has the remote_monitoring_agent built-in role. Alternatively, if it’s available in your environment, use the remote_monitoring_user built-in user.
  
  ::::{tip} If you’re using index lifecycle management, the remote monitoring user requires additional privileges to create and read indices. For more information, see Grant users access to secured resources. ::::
2. Add the username and password settings to the {{es}} output information in the {{metricbeat}} configuration file.
For more information about these configuration options, see Configure the {{es}} output.
Start {{metricbeat}} to begin collecting monitoring data.
View the monitoring data in {{kib}}.

Monitoring Metricbeat Collection

Use {{metricbeat}} to send monitoring data [monitoring-metricbeat-collection]

Configure the shipper you want to monitor [configure-shipper]

Install and configure {{metricbeat}} to collect monitoring data [configure-metricbeat]