CIS 加固指南
本文档为加固 K3s 的生产安装提供了规范性指导。它概述了解决信息安全中心（CIS）的 Kubernetes 基准控制所需的配置和控制。
K3s 已经应用了一些安全缓解措施，并在默认情况下开启，无需修改即可通过 Kubernetes CIS 的一些控制。有一些明显的例外，需要手动执行才能完全符合 CIS 基准：
- K3s 不会修改主机操作系统。任何主机级别的修改都需要用户手动修改。
- 某些针对 PodSecurityPolicies 和 NetworkPolicies 的 CIS 策略控制将限制该集群的功能。您必须选择让 K3s 配置这些，通过在命令行标志或配置文件中添加适当的选项（启用 admission plugins）以及手动应用适当的策略。在下面的章节中会有进一步的详细说明。
CIS 基准的第一节（1.1）主要涉及 pod manifest 权限和所有权。K3s 并没有将这些内容用于核心组件，因为所有的东西都被打包成一个二进制文件。
Host-level Requirements#
There are two areas of host-level requirements: kernel parameters and etcd process/directory configuration. These are outlined in this section.
Ensure protect-kernel-defaults is set#
This is a kubelet flag that will cause the kubelet to exit if the required kernel parameters are unset or are set to values that are different from the kubelet's defaults.
Note:
protect-kernel-defaultsis exposed as a top-level flag for K3s.
Set kernel parameters#
Create a file called /etc/sysctl.d/90-kubelet.conf and add the snippet below. Then run sysctl -p /etc/sysctl.d/90-kubelet.conf.
Kubernetes Runtime Requirements#
The runtime requirements to comply with the CIS Benchmark are centered around pod security (PSPs) and network policies. These are outlined in this section. K3s doesn't apply any default PSPs or network policies however K3s ships with a controller that is meant to apply a given set of network policies. By default, K3s runs with the "NodeRestriction" admission controller. To enable PSPs, add the following to the K3s start command: --kube-apiserver-arg="enable-admission-plugins=NodeRestriction,PodSecurityPolicy,ServiceAccount". This will have the effect of maintaining the "NodeRestriction" plugin as well as enabling the "PodSecurityPolicy".
PodSecurityPolicies#
When PSPs are enabled, a policy can be applied to satisfy the necessary controls described in section 5.2 of the CIS Benchmark.
Here's an example of a compliant PSP.
Before the above PSP to be effective, we need to create a couple ClusterRoles and ClusterRole. We also need to include a "system unrestricted policy" which is needed for system-level pods that require additional privileges.
These can be combined with the PSP yaml above and NetworkPolicy yaml below into a single file and placed in the /var/lib/rancher/k3s/server/manifests directory. Below is an example of a policy.yaml file.
Note: The Kubernetes critical additions such as CNI, DNS, and Ingress are ran as pods in the
kube-systemnamespace. Therefore, this namespace will have a policy that is less restrictive so that these components can run properly.
NetworkPolicies#
NOTE: K3s deploys kube-router for network policy enforcement. Support for this in K3s is currently experimental. CIS requires that all namespaces have a network policy applied that reasonably limits traffic into namespaces and pods.
Here's an example of a compliant network policy.
With the applied restrictions, DNS will be blocked unless purposely allowed. Below is a network policy that will allow for traffic to exist for DNS.
The metrics-server and Traefik ingress controller will be blocked by default if network policies are not created to allow access. Traefik v1 as packaged in K3s version 1.20 and below uses different labels than Traefik v2; ensure that you only use the sample yaml below that is associated with the version of Traefik present on your cluster.
Note: Operators must manage network policies as normal for additional namespaces that are created.
Known Issues#
The following are controls that K3s currently does not pass by default. Each gap will be explained, along with a note clarifying whether it can be passed through manual operator intervention, or if it will be addressed in a future release of K3s.
Control 1.2.15#
Ensure that the admission control plugin NamespaceLifecycle is set.
Rationale
Setting admission control policy to NamespaceLifecycle ensures that objects cannot be created in non-existent namespaces, and that namespaces undergoing termination are not used for creating the new objects. This is recommended to enforce the integrity of the namespace termination process and also for the availability of the newer objects.This can be remediated by passing this argument as a value to the enable-admission-plugins= and pass that to --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.16 (mentioned above)#
Ensure that the admission control plugin PodSecurityPolicy is set.
Rationale
A Pod Security Policy is a cluster-level resource that controls the actions that a pod can perform and what it has the ability to access. The PodSecurityPolicy objects define a set of conditions that a pod must run with in order to be accepted into the system. Pod Security Policies are comprised of settings and strategies that control the security features a pod has access to and hence this must be used to control pod access permissions.This can be remediated by passing this argument as a value to the enable-admission-plugins= and pass that to --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.22#
Ensure that the --audit-log-path argument is set.
Rationale
Auditing the Kubernetes API Server provides a security-relevant chronological set of records documenting the sequence of activities that have affected system by individual users, administrators or other components of the system. Even though currently, Kubernetes provides only basic audit capabilities, it should be enabled. You can enable it by setting an appropriate audit log path.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.23#
Ensure that the --audit-log-maxage argument is set to 30 or as appropriate.
Rationale
Retaining logs for at least 30 days ensures that you can go back in time and investigate or correlate any events. Set your audit log retention period to 30 days or as per your business requirements.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.24#
Ensure that the --audit-log-maxbackup argument is set to 10 or as appropriate.
Rationale
Kubernetes automatically rotates the log files. Retaining old log files ensures that you would have sufficient log data available for carrying out any investigation or correlation. For example, if you have set file size of 100 MB and the number of old log files to keep as 10, you would approximate have 1 GB of log data that you could potentially use for your analysis.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.25#
Ensure that the --audit-log-maxsize argument is set to 100 or as appropriate.
Rationale
Kubernetes automatically rotates the log files. Retaining old log files ensures that you would have sufficient log data available for carrying out any investigation or correlation. If you have set file size of 100 MB and the number of old log files to keep as 10, you would approximate have 1 GB of log data that you could potentially use for your analysis.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.26#
Ensure that the --request-timeout argument is set as appropriate.
Rationale
Setting global request timeout allows extending the API server request timeout limit to a duration appropriate to the user's connection speed. By default, it is set to 60 seconds which might be problematic on slower connections making cluster resources inaccessible once the data volume for requests exceeds what can be transmitted in 60 seconds. But, setting this timeout limit to be too large can exhaust the API server resources making it prone to Denial-of-Service attack. Hence, it is recommended to set this limit as appropriate and change the default limit of 60 seconds only if needed.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.27#
Ensure that the --service-account-lookup argument is set to true.
Rationale
If `--service-account-lookup` is not enabled, the apiserver only verifies that the authentication token is valid, and does not validate that the service account token mentioned in the request is actually present in etcd. This allows using a service account token even after the corresponding service account is deleted. This is an example of time of check to time of use security issue.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 1.2.33#
Ensure that the --encryption-provider-config argument is set as appropriate.
Rationale
Where `etcd` encryption is used, it is important to ensure that the appropriate set of encryption providers is used. Currently, the aescbc, kms and secretbox are likely to be appropriate options.Control 1.2.34#
Ensure that encryption providers are appropriately configured.
Rationale
`etcd` is a highly available key-value store used by Kubernetes deployments for persistent storage of all of its REST API objects. These objects are sensitive in nature and should be encrypted at rest to avoid any disclosures.This can be remediated by passing a valid configuration to k3s as outlined above.
Control 1.3.1#
Ensure that the --terminated-pod-gc-threshold argument is set as appropriate.
Rationale
Garbage collection is important to ensure sufficient resource availability and avoiding degraded performance and availability. In the worst case, the system might crash or just be unusable for a long period of time. The current setting for garbage collection is 12,500 terminated pods which might be too high for your system to sustain. Based on your system resources and tests, choose an appropriate threshold value to activate garbage collection.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 3.2.1#
Ensure that a minimal audit policy is created (Scored)
Rationale
Logging is an important detective control for all systems, to detect potential unauthorized access.This can be remediated by passing controls 1.2.22 - 1.2.25 and verifying their efficacy.
Control 4.2.7#
Ensure that the --make-iptables-util-chains argument is set to true.
Rationale
Kubelets can automatically manage the required changes to iptables based on how you choose your networking options for the pods. It is recommended to let kubelets manage the changes to iptables. This ensures that the iptables configuration remains in sync with pods networking configuration. Manually configuring iptables with dynamic pod network configuration changes might hamper the communication between pods/containers and to the outside world. You might have iptables rules too restrictive or too open.This can be remediated by passing this argument as a value to the --kube-apiserver-arg= argument to k3s server. An example can be found below.
Control 5.1.5#
Ensure that default service accounts are not actively used. (Scored)
Rationale
Kubernetes provides a default service account which is used by cluster workloads where no specific service account is assigned to the pod.
Where access to the Kubernetes API from a pod is required, a specific service account should be created for that pod, and rights granted to that service account.
The default service account should be configured such that it does not provide a service account token and does not have any explicit rights assignments.
The remediation for this is to update the automountServiceAccountToken field to false for the default service account in each namespace.
For default service accounts in the built-in namespaces (kube-system, kube-public, kube-node-lease, and default), K3s does not automatically do this. You can manually update this field on these service accounts to pass the control.
Control Plane Execution and Arguments#
Listed below are the K3s control plane components and the arguments they're given at start, by default. Commented to their right is the CIS 1.5 control that they satisfy.
The command below is an example of how the outlined remediations can be applied.
Conclusion#
If you have followed this guide, your K3s cluster will be configured to comply with the CIS Kubernetes Benchmark. You can review the CIS Benchmark Self-Assessment Guide to understand the expectations of each of the benchmarks and how you can do the same on your cluster.