Subscribe to raw Carbon Black event feed and forward to another system, such as Splunk, etc.
Clone or download
Latest commit c8d9b5d Aug 9, 2018

Cb Response Event Forwarder


The Cb Response Event Forwarder is a standalone service that will listen on the Cb Response enterprise bus and export events (both watchlist/feed hits as well as raw endpoint events, if configured) in a normalized JSON or LEEF format. The events can be saved to a file, delivered to a network service or archived automatically to an Amazon AWS S3 bucket. These events can be consumed by any external system that accepts JSON or LEEF, including Splunk and IBM QRadar.


The 4.0 Cb Response Event Forwarder has a three part processing pipeline which is specified: in YAML format - with two crucicial parts - input and output.

  1. input - this section defines a number of consumers each reading from a specific CbR server messaging bus The list of events to collect is configurable. By default all feed and watchlist hits, alerts, binary notifications, and raw sensor events are exported into JSON. The
        cb_server_url: https://zestep-centos-cbresponseserver 
        # rabbitmq options are opitional, /etc/cb/.conf is used by default
        # if the folling are ommited
        rabbit_mq_username: cb
        rabbit_mq_hostname: localhost

#Use this boolean option to control the use of the raw sensor exchange #defaults to false use_raw_exchange : true

Subscribed Events

The 4.0 event forwarder subcribes to all-events by default though the user can specify explicit options for which events to listen to as so: (within each individual input element) event_map: events_watchlist: - watchlist.# events_feed: - feed.# events_alert: - alert.# events_raw_sensor: - ingress.event.process - ingress.event.procstart - ingress.event.netconn - ingerss.event.procend - ingress.event.childproc - ingress.event.moduleload - ingress.event.module - ingress.event.filemod - ingress.event.regmod - ingress.event.tamper - ingress.event.crossprocopen - ingress.event.remotethread - ingress.event.processblock - ingress.event.emetmitigation events_binary_observed: - binaryinfo.# events_binary_upload:
- binarystore.# events_storage_parition: - events.partition.#

AMQPs tls can be configured by providing a tls: stanza within an individual input:

    client_cert: yourcert
    client_key: yourkey
    ca_cert: cacert
    tls_insecure: true # defaults false 

There are optional post processing arguments to enhance messages w/ api callbacks (currently just report information)

        #slightly different than other TLS stanzas found in input/output
        verify: false # defaults to true
        tls_12_only: false # defaults to true 
        client_cert: client.cert
        client_key: client.key
        ca_cert: ca.cert
    api_proxy_url: proxyurl
    api_token: apitoken

Just specify the post processing options within an individual input element.

  1. filter - this section of the pipeline defines an optional template for keeping/droping messages based on their contents. This is a top-level option, along side input: and output:
    template: >-
              {{if (eq .type "alert.watchlist.hit.query.binary") -}}
              {{- else -}}
              {{- end}}
  1. output - this section defines a number of outputs, which can be of differing types (file,socket, etc) and formats (json,LEEF, custom)
    - file:
        path: "output-json.txt"
            type: json 
    - file:
        path: "output-leef.txt"
            type: leef 
    - file:
        path: "output-yaml.txt"
            type: template 
            template: "{{YamlFormat .}}"

Supported Output Types in 4.0

The 4.0.0 CbR event-forwarder supports the same output options as 3.X: Each output must provide path/connection information, other configuration options specific to that output type and a format (like the old output_format) like so:

output formats

    type: json # leef, cef, template are the possible options

output options

    path: path/to/your/desiredoutput.type
        type: json
    destination: https://myserver:51337
    # overide the default template ... http_post_template: 
    #upload_empty_files: false    
    #bundle_size_max: "50000000"
    #bundle_send_timeout: "500"
    #comma_seperate_events: false 
    # defaults to false 
    #   header: headervalue
        #tls options
        type: json
   connection: tpc://host-or-ip:port 
        type: json
    connection: tcp://host-or-ip:port 


The pre-built RPM is supported via our User eXchange (Jive) and via email to

Raw Sensor Events

We have seen a performance impact when exporting all raw sensor events onto the enterprise bus. We do not recommend exporting all the events. The performance impacts are seen when the events are broadcast on the bus, by enabling the "DatastoreBroadcastEventTypes". We recommend that at most, only process and netconn events be broadcast on the event bus.

Quickstart Guide

The cb-event-forwarder can be installed on any 64-bit Linux machine running CentOS 6.x. It can be installed on the same machine as the Cb Response server, or another machine. If you are forwarding a large volume of events to QRadar (for example, all file modifications and/or registry modifications), or are forwarding events from a Cb Response cluster, then installing it on a separate machine is recommended. Otherwise, it is acceptable to install the cb-event-forwarder on the Cb Response server itself.


To install and configure the cb-event-forwarder, perform these steps as "root" on your target Linux system.

  1. Install the CbOpenSource repository if it isn't already present:

    cd /etc/yum.repos.d
    curl -O
  2. Install the RPM via YUM:

    yum install cb-event-forwarder

Configure the cb-event-forwarder

Note: The 4.X event forwarder uses Yaml files for configuration. Values will be interpreted as their literal types, rather than using 0,1,T/F to represent booleans true or false should be used.

  1. If installing on a machine other than the Cb Response server, copy the RabbitMQ username and password into the rabbit_mq_username and rabbit_mq_password variables in /etc/cb/integrations/event-forwarder/cb-event-forwarder.conf file. Also fill out the cb_server_hostname with the hostname or IP address where the Cb Response server can be reached. If the cb-event-forwarder is forwarding events from a Cb Response cluster, the cb_server_hostname should be set to the hostname or IP address of the Cb Response master node.

  2. Ensure that the configuration is valid by running the cb-event-forwarder in Check mode: /usr/share/cb/integrations/event-forwarder/cb-event-forwarder -check as root. If everything is OK, you will see a message starting with "Initialized output”. If there are any errors, those errors will be printed to your screen.

Configure Cb Response

By default, Cb publishes the feed.* and watchlist.* events over the bus (see the Events documentation for more information). If you want to capture raw sensor events or the binaryinfo.* notifications, you have to enable those features in /etc/cb/cb.conf:

  • If you are capturing raw sensor events then you also need to edit the DatastoreBroadcastEventTypes option in /etc/cb/cb.conf to enable broadcast of the raw sensor events you wish to export.
  • If you are capturing binary observed events you also need to edit the EnableSolrBinaryInfoNotifications option in /etc/cb/cb.conf and set it to true.

Cb Response needs to be restarted if any variables were changed in /etc/cb/cb.conf by executing service cb-enterprise restart.

If you are configuring the cb-event-forwarder on a Cb Response cluster, the DatastoreBroadcastEventTypes and/or EnableSolrBinaryInfoNotifications settings must be distributed to the /etc/cb/cb.conf configuration file on all minion nodes and the cluster stopped and started using the /usr/share/cb/cbcluster stop && /usr/share/cb/cbcluster start command.

Starting and Stopping the Service

Once the service is installed, it is managed by the Upstart init system in CentOS 6.x. You can control the service via the initctl command.

  • To start the service, initctl start cb-event-forwarder
  • To stop the service, initctl stop cb-event-forwarder

Once the service is installed, it is configured to start automatically on system boot.


The Cb Response event forwarder can be used to export Cb Response events in a way easily configured for Splunk. You'll need to install and configure the Splunk TA to consume the Cb Response event data. It is recommended that the event bridge use a file based output with Splunk universal forwarder configured to monitor the file.

More information about configuring the Splunk TA can be found here


The Cb Response event forwarder can forward Cb Response events in the LEEF format to QRadar. To forward Cb Response events to a QRadar server:

  1. Modify /etc/cb/integrations/event-forwarder/cb-event-forwarder.conf to include udpout: "<qradaripaddress>:<port>" (NOTE: Port is usually 514)
  2. Change the output format to LEEF in the configuration file: output_format: leef.
  3. Change the output type to UDP in the configuration file: output_type: udp.

For more information on the LEEF format, see the Events documentation.

Logging & Diagnostics

The connector logs to the directory /var/log/cb/integrations/cb-event-forwarder. An example of a successful startup log:

2015/12/07 12:57:26 cb-event-forwarder version 4.0.0 starting
2015/12/07 12:57:26 Interface address
2015/12/07 12:57:26 Interface address fe80::20c:29ff:fe85:bcd0
2015/12/07 12:57:26 Configured to capture events: [watchlist.hit.# feed.ingress.hit.# feed.query.hit.# alert.watchlist.hit.# ingress.event.process ingress.event.procstart 
ingress.event.netconn ingress.event.procend ingress.event.childproc ingress.event.moduleload 
ingress.event.module ingress.event.filemod ingress.event.regmod binaryinfo.# binarystore.file.added]
2015/12/07 12:57:26 Initialized output: File /var/cb/data/event_bridge_output.json
2015/12/07 12:57:26 Diagnostics available via HTTP at http://cbtest:33706/debug/vars
2015/12/07 12:57:26 Starting AMQP loop
2015/12/07 12:57:26 Connecting to message bus...
2015/12/07 12:57:26 Subscribed to watchlist.hit.#
2015/12/07 12:57:26 Subscribed to
2015/12/07 12:57:26 Subscribed to feed.ingress.hit.#
2015/12/07 12:57:26 Subscribed to
2015/12/07 12:57:26 Subscribed to feed.query.hit.#
2015/12/07 12:57:26 Subscribed to alert.watchlist.hit.#
2015/12/07 12:57:26 Subscribed to ingress.event.process
2015/12/07 12:57:26 Subscribed to ingress.event.procstart
2015/12/07 12:57:26 Subscribed to ingress.event.netconn
2015/12/07 12:57:26 Subscribed to ingress.event.procend
2015/12/07 12:57:26 Subscribed to ingress.event.childproc
2015/12/07 12:57:26 Subscribed to ingress.event.moduleload
2015/12/07 12:57:26 Subscribed to ingress.event.module
2015/12/07 12:57:26 Subscribed to ingress.event.filemod
2015/12/07 12:57:26 Subscribed to ingress.event.regmod
2015/12/07 12:57:26 Subscribed to binaryinfo.#
2015/12/07 12:57:26 Subscribed to binarystore.file.added
2015/12/07 12:57:26 Starting 4 message processors

In addition to the log file, the service starts an HTTP service for monitoring and debugging. The URL is available in the log file (see the “Diagnostics available” line above). The port is configurable through the http_server_port option in the configuration file.

The diagnostics are presented as a JSON formatted string. The diagnostics include operational information on the service itself, how long the service has been running, errors, and basic configuration information. An example output from the JSON status is shown below:

  "version": "3.0.0",
  "uptime": 145.617786697,
  "cmdline": [
  "connection_status": {
    "uptime": 145.471995845,
    "last_error_time": "0001-01-01T00:00:00Z",
    "last_error_text": "",
    "last_connect_time": "2015-12-08T00:22:56.566600876-05:00",
    "connected": true
  "error_count": 0,
  "input_event_count": 29,
  "memstats": {...},
  "output_event_count": 29,
  "output_status": {
    "type": "file",
    "format": "json",
    "file:/var/cb/data/event_bridge_output.json": {
      "file_name": "/var/cb/data/event_bridge_output.json",
      "last_open_time": "2015-12-08T00:22:56.430385291-05:00"
  "subscribed_events": [

Building from source

It is recommended to use the latest avialable golang toolchain for your environment

Go-Plugin support is crucial to the plugin implementation and is subject to bugs in MacOSX, therefore plugins can not be at present supported on that system. (Although the forwarder as a whole, and the standard outoputs will work just fine)

Setup your GOPATH environment variable. See for details

go get
go get -u
go get -u
go generate ./...
go get ./...
go build