System Moduleedit

This functionality is in beta and is subject to change. The design and code is less mature than official GA features and is being provided as-is with no warranties. Beta features are not subject to the support SLA of official GA features.

The system module collects various security related information about a system. All datasets send both periodic state information (e.g. all currently running processes) and real-time changes (e.g. when a new process starts or stops).

The module is fully implemented for Linux on x86. Currently, the socket module is not available on ARM. Some datasets are also available for macOS (Darwin) and Windows.

How it worksedit

Each dataset sends two kinds of information: state and events.

State information is sent periodically and (for some datasets) on startup. A state update will consist of one event per object that is currently active on the system (e.g. a process). All events belonging to the same state update will share the same UUID in

The frequency of state updates can be controlled for all datasets using the state.period configuration option. Overrides are available per dataset. The default is 12h.

Event information is sent as the events occur (e.g. a process starts or stops). All datasets are currently using a poll model to retrieve their data. The frequency of these polls is controlled by the period configuration parameter.

Entity IDsedit

This module populates entity_id fields to uniquely identify entities (users, packages, processes…​) within a host. This requires Auditbeat to obtain a unique identifier for the host:

  • Windows: Uses the HKLM\Software\Microsoft\Cryptography\MachineGuid registry key.
  • macOS: Uses the value returned by gethostuuid(2) system call.
  • Linux: Uses the content of one of the following files, created by either systemd or dbus:

    • /etc/machine-id
    • /var/lib/dbus/machine-id
    • /var/db/dbus/machine-id

Under CentOS 6.x, it’s possible that none of the files above exist. In that case, running dbus-uuidgen --ensure (provided by the dbus package) will generate one for you.

Example dashboardedit

The module comes with a sample dashboard:

Auditbeat System Overview Dashboard

Configuration optionsedit

This module has some configuration options for controlling its behavior. The following example shows all configuration options with their default values for Linux.

It is recommended to configure some datasets separately. See below for a sample suggested configuration.

- module: system
    - host
    - login
    - package
    - process
    - socket
    - user
  period: 10s
  state.period: 12h

  socket.include_localhost: false

  user.detect_password_changes: true

This module also supports the standard configuration options described later.

The interval at which the datasets send full state information. This option can be overridden per dataset using {dataset}.state.period.
If the user dataset is configured and this option is set to true, Auditbeat will read password information in /etc/passwd and /etc/shadow to detect password changes. A hash will be kept locally in the beat.db file to detect changes between Auditbeat restarts. The beat.db file should be readable only by the root user and be treated similar to the shadow file itself.

Standard configuration optionsedit

You can specify the following options for any Auditbeat module.

The name of the module to run.
A list of datasets to execute.
A Boolean value that specifies whether the module is enabled.
The frequency at which the datasets check for changes. If a system is not reachable, Auditbeat returns an error for each period. This setting is required. For most datasets, especially process and socket, a shorter period is recommended.
A dictionary of fields that will be sent with the dataset event. This setting is optional.
A list of tags that will be sent with the dataset event. This setting is optional.

A list of processors to apply to the data generated by the dataset.

See Processors for information about specifying processors in your config.


If present, this formatted string overrides the index for events from this module (for elasticsearch outputs), or sets the raw_index field of the event’s metadata (for other outputs). This string can only refer to the agent name and version and the event timestamp; for access to dynamic fields, use output.elasticsearch.index or a processor.

Example value: "%{[]}-myindex-%{+yyyy.MM.dd}" might expand to "auditbeat-myindex-2019.12.13".

If this option is set to true, fields with null values will be published in the output document. By default, keep_null is set to false.
A name given by the user to the service the data is collected from. It can be used for example to identify information collected from nodes of different clusters with the same service.type.

Suggested configurationedit

Processes and sockets can be short-lived, so the chance of missing an update increases if the polling interval is too large.

On the other hand, host and user information is unlikely to change frequently, so a longer polling interval can be used.

- module: system
    - host
    - login
    - package
    - user
  period: 1m

  user.detect_password_changes: true

- module: system
    - process
    - socket
  period: 1s

Example configurationedit

The System module supports the common configuration options that are described under configuring Auditbeat. Here is an example configuration:

- module: system
    - package # Installed, updated, and removed packages

  period: 2m # The frequency at which the datasets check for changes

- module: system
    - host    # General host information, e.g. uptime, IPs
    - login   # User logins, logouts, and system boots.
    - process # Started and stopped processes
    - socket  # Opened and closed sockets
    - user    # User information

  # How often datasets send state updates with the
  # current state of the system (e.g. all currently
  # running processes, all open sockets).
  state.period: 12h

  # Enabled by default. Auditbeat will read password fields in
  # /etc/passwd and /etc/shadow and store a hash locally to
  # detect any changes.
  user.detect_password_changes: true

  # File patterns of the login record files.
  login.wtmp_file_pattern: /var/log/wtmp*
  login.btmp_file_pattern: /var/log/btmp*


The following datasets are available: