- Remove "Overview" header - Cut "MDM required" and "Fleetd required" sections - Move "How to add" above "Requirements." - Cut the "Audit vs. remediation" section
5.3 KiB
CIS Benchmarks
Available in Fleet Premium.
CIS Benchmarks represent the consensus-based effort of cybersecurity experts globally to help you protect your systems against threats more confidently. For more information about CIS Benchmarks check out Center for Internet Security's website.
Fleet has implemented native support for CIS Benchmarks for the following platforms:
- macOS 13.0 Ventura
- macOS 14.0 Sonoma
- Windows 10 Enterprise
- Windows 11 Enterprise
Where possible, each CIS Benchmark is implemented with a policy query in Fleet.
These benchmarks are intended to gauge your organization's security posture, rather than the current state of a given host. A host may fail a CIS Benchmark policy despite having the correct settings enabled if there is no configuration profile or Group Policy Object (GPO) in place to enforce the setting. For example, this is the query for CIS - Ensure FileVault Is Enabled (MDM Required):
SELECT 1 WHERE
EXISTS (
SELECT 1 FROM managed_policies WHERE
domain='com.apple.MCX' AND
name='dontAllowFDEDisable' AND
(value = 1 OR value = 'true') AND
username = ''
)
AND NOT EXISTS (
SELECT 1 FROM managed_policies WHERE
domain='com.apple.MCX' AND
name='dontAllowFDEDisable' AND
(value != 1 AND value != 'true')
)
AND EXISTS (
SELECT 1 FROM disk_encryption WHERE
user_uuid IS NOT "" AND
filevault_status = 'on'
);
Two things are being evaluated in this policy:
- Is FileVault currently enabled?
- Is there a profile in place that prevents FileVault from being disabled?
If either of these conditions fails, the host is considered to be failing the policy.
How to add CIS Benchmarks
All CIS policies are stored under our restricted licensed folder ee/cis/.
How to import them to Fleet:
# Download policy queries from Fleet's repository
# macOS 13
wget https://raw.githubusercontent.com/fleetdm/fleet/main/ee/cis/macos-13/cis-policy-queries.yml
# Windows 10 (note the same file name. Rename as needed.)
wget https://raw.githubusercontent.com/fleetdm/fleet/main/ee/cis/win-10/cis-policy-queries.yml
# Windows 11 (note the same file name. Rename as needed.)
wget https://raw.githubusercontent.com/fleetdm/fleet/main/ee/cis/win-11/cis-policy-queries.yml
# Apply the downloaded policies to Fleet for all files.
fleetctl apply --context <context> -f <path-to-macOS-13-policies> --policies-team <team-name>
fleetctl apply --context <context> -f <path-to-windows-10-policies> --policies-team <team-name>
fleetctl apply --context <context> -f <path-to-windows-11-policies> --policies-team <team-name>
To apply the policies on a specific team use the --policies-team flag:
fleetctl apply --policies-team "Workstations" -f cis-policy-queries.yml
Levels 1 and 2
CIS designates various benchmarks as Level 1 or Level 2 to describe the level of thoroughness and burden that each benchmark represents.
Each benchmark is tagged as CIS_Level1 or CIS_Level2.
Level 1
Items in this profile intend to:
- be practical and prudent;
- provide a clear security benefit; and
- not inhibit the utility of the technology beyond acceptable means.
Level 2
This profile extends the "Level 1" profile. Items in this profile exhibit one or more of the following characteristics:
- are intended for environments or use cases where security is paramount or acts as defense in depth measure
- may negatively inhibit the utility or performance of the technology.
Requirements
Following are the requirements to use the CIS Benchmarks in Fleet:
- Devices must be running
fleetd, Fleet's lightweight agent. - Some CIS Benchmarks explicitly involve verifying MDM-based controls, so devices must be enrolled to an MDM solution.
- On macOS, the orbit component of fleetd must have "Full Disk Access", see Grant Full Disk Access to Osquery on macOS.
Limitations
Certain benchmarks cannot be automated by a policy in Fleet. For a list of specific benchmarks which are not covered, please visit the README for each benchmark:
Performance testing
In August 2023, we completed scale testing on 10k Windows hosts and 70k macOS hosts. Ultimately, we validated both server and host performance at that scale.
Detailed results are here.