e6dbdf3855
docs for https://github.com/fleetdm/fleet/issues/13304 and it's sibling PR: https://github.com/fleetdm/fleet/pull/13564
23 KiB
Enroll hosts
- Enroll hosts
Introduction
Fleet gathers information from an osquery agent installed on each of your hosts. The recommended way to install osquery is using Fleetd, which includes Orbit, Fleet's lightweight osquery runtime and auto-updater.
You can also install plain osquery on your hosts and connect to Fleet using osquery's TLS API plugins.
For ChromeOS hosts, the fleetd Chrome extension is installed instead of osquery.
Enroll hosts with Fleetd
To create a Fleet installer, you can use the fleetctl package command. To use the fleetctl package command, you must first install the fleetctl command-line tool. Learn how to install fleetctl.
The fleetctl package can create a Fleetd installer, which adds macOS hosts (.pkg), Windows hosts (.msi), or Linux hosts (.deb or .rpm) to Fleet.
The following command creates a Fleetd installer, .pkg file, which adds macOS hosts to Fleet. Locate this osquery installer in the folder where the fleetctl package command is run.
fleetctl package --type pkg --fleet-url=[YOUR FLEET URL] --enroll-secret=[YOUR ENROLLMENT SECRET]
Note: The only configuration option required to create an installer is
--type. However, to communicate with a Fleet instance, you'll need to specify a--fleet-urland--enroll-secret. When those flags are not set, fleetd will log errors.
When you install the generated Fleetd installer on a host, this host will automatically enroll in the specified Fleet instance.
Signing installers
Note: Currently, the
fleetctl packagedoes not provide support for signing Windows Fleetd installers. Windows installers can be signed after building.
The fleetctl package provides support for signing and notarizing macOS osquery installers via the
--sign-identity and --notarize flags.
Check out the example below:
AC_USERNAME=appleid@example.com AC_PASSWORD=app-specific-password fleetctl package --type pkg --sign-identity=[PATH TO SIGN IDENTITY] --notarize --fleet-url=[YOUR FLEET URL] --enroll-secret=[YOUR ENROLLMENT SECRET]
The above command should be run on a macOS device, as the notarizing and signing of macOS Fleetd installers can only be done on macOS devices.
Also, remember to replace both AC_USERNAME and AC_PASSWORD environment variables with your Apple ID and a valid app-specific password, respectively. Some organizations (notably those with Apple Enterprise Developer Accounts) may also need to specify AC_TEAM_ID. This value can be found on the Apple Developer "Membership" page under "Team ID."
Including Fleet Desktop
Fleet Desktop requires a Fleet version of 4.12.0 and above. To check your Fleet version, select the avatar on the right side of the top bar and select My account. Your Fleet version is displayed below the Get API token button.
Hosts without Fleet Desktop currently installed require a new installer to be generated and run on the target host.
How to generate an installer that includes Fleet Desktop in the Fleet UI:
- On the top bar in the Fleet UI, select Hosts > Add hosts.
- Select the Include Fleet Desktop checkbox.
- Select the clipboard icon to copy the
fleetctl packagecommand. - In your terminal application, paste and run the copied command.
Alternatively, you can generate an installer that includes Fleet Desktop in fleetctl package by appending the --fleet-desktop flag.
Fleet Desktop is supported on macOS, Windows, and Linux. Check out the supported Linux distributions and versions here on GitHub.
Once installed on the target host, Fleet Desktop will be managed by Fleetd. To learn more about Fleetd updates, see here.
To prevent this auto-update behavior, you can turn off auto-updates via the --disable-updates flag or you can set a specific channel using the --desktop-channel flag.
Adding multiple hosts
If you're managing an enterprise environment with multiple hosts, you likely have an enterprise deployment tool like Munki, Jamf Pro, Chef, Ansible, or Puppet to deliver software to your hosts.
You can distribute your Fleetd installer and add all your hosts to Fleet using your software management tool of choice.
Automatically adding hosts to a team
Applies only to Fleet Premium
ℹ️ Fleet 4.0 introduced Teams.
The Teams feature in Fleet allows you to place hosts in exclusive groups. With hosts segmented into Teams, you can apply unique queries and give users access to only the hosts in specific Teams.
You can add a host to a team by generating and using a unique osquery installer for a team or by manually transferring a host to a team in the Fleet UI.
To generate an osquery installer for a team:
- First, create a team in Fleet by selecting Create team in Settings > Teams.
- Then, navigate to Hosts and select your team.
- Next, select Add hosts and copy the
fleetctl packagecommand for the platform (macOS, Windows, Linux) of the hosts you'd like to add to a team in Fleet. - Run the copied
fleetctl packagecommand and distribute your installer to add your hosts to a team in Fleet.
Configuration options
The following command-line flags allow you to configure an osquery installer further to communicate with a specific Fleet instance.
| Flag | Options |
|---|---|
| --type | Required - Type of package to build. Options: pkg(macOS),msi(Windows), deb(Debian based Linux), rpm(RHEL, CentOS, etc.) |
| --fleet-desktop | Include Fleet Desktop. |
| --enroll-secret | Enroll secret for authenticating to Fleet server |
| --fleet-url | URL (host:port) of Fleet server |
| --fleet-certificate | Path to server certificate bundle |
| --identifier | Identifier for package product (default: com.fleetdm.orbit) |
| --version | Version for package product (default: 0.0.3) |
| --insecure | Disable TLS certificate verification (default: false) |
| --service | Install osquery with a persistence service (launchd, systemd, etc.) (default: true) |
| --sign-identity | Identity to use for macOS codesigning |
| --notarize | Whether to notarize macOS packages (default: false) |
| --disable-updates | Disable auto updates on the generated package (default: false) |
| --osqueryd-channel | Update channel of osqueryd to use (default: stable) |
| --orbit-channel | Update channel of Orbit to use (default: stable) |
| --desktop-channel | Update channel of desktop to use (default: stable) |
| --update-url | URL for update server (default: https://tuf.fleetctl.com) |
| --update-roots | Root key JSON metadata for update server (from fleetctl updates roots) |
| --use-system-configuration | Try to read --fleet-url and --enroll-secret using configuration in the host (currently only macOS profiles are supported) |
| --enable-scripts | Enable script execution (default: false) |
| --debug | Enable debug logging (default: false) |
| --verbose | Log detailed information when building the package (default: false) |
| --help, -h | show help (default: false) |
Fleet supports other methods for adding your hosts to Fleet, such as the plain osquery binaries or Kolide Osquery Launcher.
Add hosts with plain osquery
Osquery's TLS API plugin lets you use the native osqueryd binaries to connect to Fleet.
You can find various ways to install osquery on your hosts at https://osquery.io/downloads. Once you have installed osquery, you need to do three things on your hosts:
- Set up your Fleet enroll secret.
- Provide the TLS certificate that osquery will use to communicate with Fleet.
- Configure and launch osqueryd.
Set up your Fleet enroll secret
The enroll secret is a value that osquery provides to authenticate with Fleet. There are a few ways you can set the enroll secret on the hosts which you control. You can either set the value as
- a value of an environment variable (a common name is
OSQUERY_ENROLL_SECRET) - the content of a local file (a common path is
/etc/osquery/enroll_secret)
The value of the environment variable or content of the file should be a secret shared between the osqueryd client and the Fleet server. This is osqueryd's passphrase which it uses to authenticate with Fleet, convincing Fleet that it is actually one of your hosts. The passphrase could be whatever you'd like, but it would be prudent to have the passphrase long, complex, mixed-case, etc. When you launch the Fleet server, you should specify this same value.
If you use an environment variable for this, you can specify it with the --enroll_secret_env flag when you launch osqueryd. If you use a local file for this, you can specify its path with the --enroll_secret_path flag.
To retrieve the enroll secret, use the "Add New Host" dialog in the Fleet UI or
fleetctl get enroll_secret).
If your organization has a robust internal public key infrastructure (PKI) and you already deploy TLS client certificates to each host to uniquely identify them, then osquery supports an advanced authentication mechanism that takes advantage of this. Fleet can be fronted with a proxy that will perform the TLS client authentication.
Provide the TLS certificate that osquery will use to communicate with Fleet
When Fleet uses a self-signed certificate, osquery agents will need a copy of that certificate in order to authenticate the Fleet server. If clients connect directly to the Fleet server, you can download the certificate through the Fleet UI. From the main dashboard (/hosts/manage), click Add New Host and Fetch Certificate. If Fleet is running behind a load-balancer that terminates TLS, you will have to talk to your system administrator about where to find this certificate.
It is important that the CN of this certificate matches the hostname or IP that osqueryd clients will use to connect.
Specify the path to this certificate with the --tls_server_certs flag when you launch osqueryd.
Configure and launch osquery
In order for osquery to connect to the fleet server, there are some flags that need to be set:
--enroll_secret_path=/etc/osquery/enroll_secret
--tls_server_certs=/etc/osquery/fleet.crt
--tls_hostname=fleet.example.com
--host_identifier=uuid
--enroll_tls_endpoint=/api/osquery/enroll
--config_plugin=tls
--config_tls_endpoint=/api/osquery/config
--config_refresh=10
--disable_distributed=false
--distributed_plugin=tls
--distributed_interval=10
--distributed_tls_max_attempts=3
--distributed_tls_read_endpoint=/api/osquery/distributed/read
--distributed_tls_write_endpoint=/api/osquery/distributed/write
--logger_plugin=tls
--logger_tls_endpoint=/api/osquery/log
--logger_tls_period=10
These can be specified directly in the command line or saved to a flag file.
Launching osqueryd using command-line flags
Assuming that you are deploying your enroll secret in the file /etc/osquery/enroll_secret and your osquery server certificate is at /etc/osquery/fleet.crt, you could copy and paste the following command with the following flags (be sure to replace fleet.acme.net with the hostname or IP of your Fleet installation):
sudo osqueryd \
--enroll_secret_path=/etc/osquery/enroll_secret \
--tls_server_certs=/etc/osquery/fleet.crt \
--tls_hostname=fleet.example.com \
--host_identifier=uuid \
--enroll_tls_endpoint=/api/osquery/enroll \
--config_plugin=tls \
--config_tls_endpoint=/api/osquery/config \
--config_refresh=10 \
--disable_distributed=false \
--distributed_plugin=tls \
--distributed_interval=10 \
--distributed_tls_max_attempts=3 \
--distributed_tls_read_endpoint=/api/osquery/distributed/read \
--distributed_tls_write_endpoint=/api/osquery/distributed/write \
--logger_plugin=tls \
--logger_tls_endpoint=/api/osquery/log \
--logger_tls_period=10
If your osquery server certificate is deployed to a path that is not /etc/osquery/fleet.crt, be sure to update the --tls_server_certs flag. Similarly, if your enroll secret is in an environment variable that is not called OSQUERY_ENROLL_SECRET. Then, be sure to update the --enroll_secret_env environment variable.
If your enroll secret is defined in a local file, specify the file's path with the --enroll_secret_path flag instead of using the --enroll_secret_env flag.
Launching osqueryd using a flag file
For your convenience, osqueryd supports putting all your flags into a single file. We suggest deploying this file to /etc/osquery/fleet.flags. If you've deployed the appropriate osquery flags to that path, you could simply launch osquery via:
osqueryd --flagfile=/etc/osquery/fleet.flags
When using a flag file on Windows, make sure that file paths in the flag file are absolute and not quoted. For example, in C:\Program Files\osquery\osquery.flags:
--tls_server_certs=C:\Program Files\osquery\fleet.pem
--enroll_secret_path=C:\Program Files\osquery\secret.txt
Migrating from plain osquery to osquery installer
The following is a strategy for migrating a plain osquery deployment. Unlike plain osquery, Fleet's osquery installer supports the automatic updating of osquery on your hosts so that you don't have to deploy a new package for every new osquery release.
Generate installer
fleetctl package --type [pkg|msi|deb|rpm] --fleet-url [fleet-hostname:port] --enroll-secret [secret]
If you currently ship a certificate (fleet.pem), also include this in the generated package with
--fleet-certificate [/path/to/fleet.pem].
Fleet automatically manages most of the osquery flags to connect to the Fleet server. There's no
need to set any of the flags mentioned above in Configure and launch osquery. To
include other osquery flags, provide a flagfile when packaging with --osquery-flagfile [/path/to/osquery.flags].
Test the installers on each platform before initiating the migration.
Migrate
Install the generated package using your standard deployment tooling (Chef, Puppet, etc.). At this time, uninstall the existing osquery.
If the existing enrolled hosts use --host_identifier=uuid (or the uuid setting for Fleet's
osquery_host_identifier), the new
installation should appear as the same host in the Fleet UI. If other settings are used, duplicate
entries will appear in the Fleet UI. The older entries can be automatically cleaned up with the host
expiration setting. To configure this setting, in the Fleet UI, head to Settings > Organization settings > Advanced options.
Add Chromebooks with the fleetd Chrome extension
The fleetd Chrome browser extension is supported on ChromeOS operating systems that are managed using Google Admin. It is not intended for non-ChromeOS hosts with the Chrome browser installed.
Visit the Google Admin console. In the navigation menu, visit Devices > Chrome > Apps & Extensions > Users & browsers.
Select the relevant organizational unit, users, or group where you want the fleetd Chrome extension to be installed.
In the bottom right, click the yellow "+" button and select "Add Chrome app or extension by ID."
Visit your Fleet instance and select Hosts > Add Hosts and select ChromeOS in the popup modal.
Enter the "Extension ID," "Installation URL," and "Policy for extensions" using the data provided in the modal.
Under "Installation Policy", select "Force install". Under "Update URL", select "Installation URL (see above)".
For the fleetd Chrome extension to have full access to Chrome data, it must be force-installed by enterprise policy as per above
Grant full disk access to osquery on macOS
macOS does not allow applications to access all system files by default. If you are using MDM, which is required to deploy these profiles, you can deploy a "Privacy Preferences Policy Control" policy to grant Fleetd or osquery that level of access. This is necessary to query for files located in protected paths as well as to use event tables that require access to the EndpointSecurity API, such as es_process_events.
Creating the configuration profile
Obtaining identifiers
If you use plain osquery, instructions are available here.
On a system with osquery installed via the Fleet osquery installer (Fleetd), obtain the
CodeRequirement of Fleetd by running:
codesign -dr - /opt/orbit/bin/orbit/macos/stable/orbit
The output should be similar or identical to:
Executable=/opt/orbit/bin/orbit/macos/edge/orbit
designated => identifier "com.fleetdm.orbit" and anchor apple generic and certificate 1[field.1.2.840.113635.100.6.2.6] /* exists */ and certificate leaf[field.1.2.840.113635.100.6.1.13] /* exists */ and certificate leaf[subject.OU] = "8VBZ3948LU"
NOTE: Depending on the version of
fleetctlused to package and install Orbit, as well as the update channel you've specified, the executable path may differ. Fleetctl versions <= 4.13.2 would install Orbit to/var/lib/orbitinstead of/opt/orbit.
Note down the executable path and the entire identifier.
Osqueryd will inherit the privileges from Orbit and does not need explicit permissions.
Creating the profile
Depending on your MDM, this might be possible in the UI or require a custom profile. If your MDM has a feature to configure Policy Preferences, follow these steps:
- Configure the identifier type to “path.”
- Paste the full path to Orbit as the identifier.
- Paste the full code signing identifier into the Code requirement field.
- Allow “Access all files.” Access to Downloads, Documents, etc., is inherited from this.
If your MDM does not have built-in support for privacy preferences profiles, you can use PPPC-Utility to create a profile with those values, then upload it to your MDM as a custom profile.
Test the profile
Link the profile to a test group that contains at least one Mac. Once the computer has received the profile, which you can verify by looking at Profiles in System Preferences, run this query from Fleet:
SELECT * FROM file WHERE path LIKE '/Users/%/Downloads/%%';
If this query returns files, the profile was applied, as Downloads is a protected location. You can now enjoy the benefits of osquery on all system files and start using the es_process_events table!
If this query does not return data, you can look at operating system logs to confirm whether or not full disk access has been applied.
See the last hour of logs related to TCC permissions with this command:
log show --predicate 'subsystem == "com.apple.TCC"' --info --last 1h
You can then look for orbit or osquery to narrow down results.