×

Installing RHACS on secured clusters by using Helm charts

You can install RHACS on secured clusters by using Helm charts with no customization, using the default values, or with customizations of configuration parameters.

Installing RHACS on secured clusters by using Helm charts without customizations

Adding the Helm chart repository

Procedure
  • Add the RHACS charts repository.

    $ helm repo add rhacs https://mirror.openshift.com/pub/rhacs/charts/

The Helm repository for Red Hat Advanced Cluster Security for Kubernetes includes Helm charts for installing different components, including:

  • Central services Helm chart (central-services) for installing the centralized components (Central and Scanner).

    You deploy centralized components only once and you can monitor multiple separate clusters by using the same installation.

  • Secured Cluster Services Helm chart (secured-cluster-services) for installing the per-cluster (Sensor and Admission controller) and per-node (Collector) components.

    Deploy the per-cluster components into each cluster that you want to monitor and deploy the per-node components in all nodes that you want to monitor.

Verification
  • Run the following command to verify the added chart repository:

    $ helm search repo -l rhacs/

Installing the secured-cluster-services Helm chart without customization

Use the following instructions to install the secured-cluster-services Helm chart to deploy the per-cluster and per-node components (Sensor, Admission controller, and Collector).

To install Collector on systems that have Unified Extensible Firmware Interface (UEFI) and that have Secure Boot enabled, you must use eBPF probes because kernel modules are unsigned, and the UEFI firmware cannot load unsigned packages. Collector identifies Secure Boot status at the start and switches to eBPF probes if required.

Prerequisites
  • You must have generated RHACS init bundle for your cluster.

  • You must have the address and the port number that you are exposing the Central service on.

Procedure
  • Run the following command on your Kubernetes based clusters:

    $ helm install -n stackrox --create-namespace \
        stackrox-secured-cluster-services rhacs/secured-cluster-services \
        -f <path_to_cluster_init_bundle.yaml> \ (1)
        --set clusterName=<name_of_the_secured_cluster> \
        --set centralEndpoint=<endpoint_of_central_service> (2)
    1 Use the -f option to specify the path for the init bundle.
    2 Specify the address and port number for Central. For example, acs.domain.com:443.
  • Run the following command on OpenShift Container Platform clusters:

    $ helm install -n stackrox --create-namespace \
        stackrox-secured-cluster-services rhacs/secured-cluster-services \
        -f <path_to_cluster_init_bundle.yaml> \ (1)
        --set clusterName=<name_of_the_secured_cluster> \
        --set centralEndpoint=<endpoint_of_central_service> (2)
        --set scanner.disable=false
    1 Use the -f option to specify the path for the init bundle.
    2 Specify the address and port number for Central. For example, acs.domain.com:443.

Configuring the secured-cluster-services Helm chart with customizations

This section describes Helm chart configuration parameters that you can use with the helm install and helm upgrade commands. You can specify these parameters by using the --set option or by creating YAML configuration files.

Create the following files for configuring the Helm chart for installing Red Hat Advanced Cluster Security for Kubernetes:

  • Public configuration file values-public.yaml: Use this file to save all non-sensitive configuration options.

  • Private configuration file values-private.yaml: Use this file to save all sensitive configuration options. Ensure that you store this file securely.

While using the secured-cluster-services Helm chart, do not modify the values.yaml file that is part of the chart.

Configuration parameters

Parameter Description

clusterName

Name of your cluster.

centralEndpoint

Address, including port number, of the Central endpoint. If you are using a non-gRPC capable load balancer, use the WebSocket protocol by prefixing the endpoint address with wss://. When configuring multiple clusters, use the hostname for the address (for example, central.example.com:443).

sensor.endpoint

Address of the Sensor endpoint including port number.

sensor.imagePullPolicy

Image pull policy for the Sensor container.

sensor.serviceTLS.cert

The internal service-to-service TLS certificate that Sensor uses.

sensor.serviceTLS.key

The internal service-to-service TLS certificate key that Sensor uses.

sensor.resources.requests.memory

The memory request for the Sensor container. Use this parameter to override the default value.

sensor.resources.requests.cpu

The CPU request for the Sensor container. Use this parameter to override the default value.

sensor.resources.limits.memory

The memory limit for the Sensor container. Use this parameter to override the default value.

sensor.resources.limits.cpu

The CPU limit for the Sensor container. Use this parameter to override the default value.

sensor.nodeSelector

Specify a node selector label as label-key: label-value to force Sensor to only schedule on nodes with the specified label.

sensor.tolerations

If the node selector selects tainted nodes, use this parameter to specify a taint toleration key, value, and effect for Sensor. This parameter is mainly used for infrastructure nodes.

image.main.name

The name of the main image.

image.collector.name

The name of the Collector image.

image.main.registry

Address of the registry you are using for the main image.

image.collector.registry

Address of the registry you are using for the Collector image.

image.main.pullPolicy

Image pull policy for main images.

image.collector.pullPolicy

Image pull policy for the Collector images.

image.main.tag

Tag of main image to use.

image.collector.tag

Tag of collector image to use.

collector.collectionMethod

Either EBPF, KERNEL_MODULE, or NO_COLLECTION.

collector.imagePullPolicy

Image pull policy for the Collector container.

collector.complianceImagePullPolicy

Image pull policy for the Compliance container.

collector.disableTaintTolerations

If you specify false, tolerations are applied to Collector, and the collector pods can schedule onto all nodes with taints. If you specify it as true, no tolerations are applied, and the collector pods are not scheduled onto nodes with taints.

collector.resources.requests.memory

The memory request for the Collector container. Use this parameter to override the default value.

collector.resources.requests.cpu

The CPU request for the Collector container. Use this parameter to override the default value.

collector.resources.limits.memory

The memory limit for the Collector container. Use this parameter to override the default value.

collector.resources.limits.cpu

The CPU limit for the Collector container. Use this parameter to override the default value.

collector.complianceResources.requests.memory

The memory request for the Compliance container. Use this parameter to override the default value.

collector.complianceResources.requests.cpu

The CPU request for the Compliance container. Use this parameter to override the default value.

collector.complianceResources.limits.memory

The memory limit for the Compliance container. Use this parameter to override the default value.

collector.complianceResources.limits.cpu

The CPU limit for the Compliance container. Use this parameter to override the default value.

collector.serviceTLS.cert

The internal service-to-service TLS certificate that Collector uses.

collector.serviceTLS.key

The internal service-to-service TLS certificate key that Collector uses.

admissionControl.listenOnCreates

This setting controls whether Kubernetes is configured to contact Red Hat Advanced Cluster Security for Kubernetes with AdmissionReview requests for workload creation events.

admissionControl.listenOnUpdates

When you set this parameter as false, Red Hat Advanced Cluster Security for Kubernetes creates the ValidatingWebhookConfiguration in a way that causes the Kubernetes API server not to send object update events. Since the volume of object updates is usually higher than the object creates, leaving this as false limits the load on the admission control service and decreases the chances of a malfunctioning admission control service.

admissionControl.listenOnEvents

This setting controls whether the cluster is configured to contact Red Hat Advanced Cluster Security for Kubernetes with AdmissionReview requests for Kubernetes exec and portforward events. Red Hat Advanced Cluster Security for Kubernetes does not support this feature on OpenShift Container Platform 3.11.

admissionControl.dynamic.enforceOnCreates

This setting controls whether Red Hat Advanced Cluster Security for Kubernetes evaluates policies; if it is disabled, all AdmissionReview requests are automatically accepted.

admissionControl.dynamic.enforceOnUpdates

This setting controls the behavior of the admission control service. You must specify listenOnUpdates as true for this to work.

admissionControl.dynamic.scanInline

If you set this option to true, the admission control service requests an image scan before making an admission decision. Since image scans take several seconds, enable this option only if you can ensure that all images used in your cluster are scanned before deployment (for example, by a CI integration during image build). This option corresponds to the Contact image scanners option in the RHACS Portal.

admissionControl.dynamic.disableBypass

Set it to true to disable bypassing the Admission controller.

admissionControl.dynamic.timeout

The maximum time, in seconds, Red Hat Advanced Cluster Security for Kubernetes should wait while evaluating admission review requests. Use this to set request timeouts when you enable image scanning. If the image scan runs longer than the specified time, Red Hat Advanced Cluster Security for Kubernetes accepts the request.

admissionControl.resources.requests.memory

The memory request for the Admission Control container. Use this parameter to override the default value.

admissionControl.resources.requests.cpu

The CPU request for the Admission Control container. Use this parameter to override the default value.

admissionControl.resources.limits.memory

The memory limit for the Admission Control container. Use this parameter to override the default value.

admissionControl.resources.limits.cpu

The CPU limit for the Admission Control container. Use this parameter to override the default value.

admissionControl.nodeSelector

Specify a node selector label as label-key: label-value to force Admission Control to only schedule on nodes with the specified label.

admissionControl.tolerations

If the node selector selects tainted nodes, use this parameter to specify a taint toleration key, value, and effect for Admission Control. This parameter is mainly used for infrastructure nodes.

admissionControl.serviceTLS.cert

The internal service-to-service TLS certificate that Admission Control uses.

admissionControl.serviceTLS.key

The internal service-to-service TLS certificate key that Admission Control uses.

registryOverride

Use this parameter to override the default docker.io registry. Specify the name of your registry if you are using some other registry.

collector.disableTaintTolerations

If you specify false, tolerations are applied to Collector, and the Collector pods can schedule onto all nodes with taints. If you specify it as true, no tolerations are applied, and the Collector pods are not scheduled onto nodes with taints.

createUpgraderServiceAccount

Specify true to create the sensor-upgrader account. By default, Red Hat Advanced Cluster Security for Kubernetes creates a service account called sensor-upgrader in each secured cluster. This account is highly privileged but is only used during upgrades. If you do not create this account, you must complete future upgrades manually if the Sensor does not have enough permissions.

createSecrets

Specify false to skip the orchestrator secret creation for the Sensor, Collector, and Admission controller.

collector.slimMode

Specify true if you want to use a slim Collector image for deploying Collector. Using slim Collector images requires Central to provide the matching eBPF probe or kernel module. If you are running Red Hat Advanced Cluster Security for Kubernetes in offline mode, you must download a kernel support package from stackrox.io and upload it to Central for slim Collectors to function. Otherwise, you must ensure that Central can access the online probe repository hosted at https://collector-modules.stackrox.io/.

sensor.resources

Resource specification for Sensor.

admissionControl.resources

Resource specification for Admission controller.

collector.resources

Resource specification for Collector.

collector.complianceResources

Resource specification for Collector’s Compliance container.

exposeMonitoring

If you set this option to true, Red Hat Advanced Cluster Security for Kubernetes exposes Prometheus metrics endpoints on port number 9090 for the Sensor, Collector, and the Admission controller.

auditLogs.disableCollection

If you set this option to true, Red Hat Advanced Cluster Security for Kubernetes disables the audit log detection features used to detect access and modifications to configuration maps and secrets.

scanner.disable

If you set this option to false, Red Hat Advanced Cluster Security for Kubernetes deploys a lightweight scanner and Scanner DB in the secured cluster to allow scanning images on OpenShift Container Registry. Enabling Scanner is only supported on OpenShift. Defaults to true

scanner.dbTolerations

If the node selector selects tainted nodes, use this parameter to specify a taint toleration key, value, and effect for Scanner DB.

scanner.replicas

Resource specification for Collector’s Compliance container.

scanner.logLevel

Setting this parameter allows you to modify the scanner log level. Use this option only for troubleshooting purposes.

scanner.autoscaling.disable

If you set this option to true, Red Hat Advanced Cluster Security for Kubernetes disables autoscaling on the Scanner deployment.

scanner.autoscaling.minReplicas

The minimum number of replicas for autoscaling. Defaults to 2.

scanner.autoscaling.maxReplicas

The maximum number of replicas for autoscaling. Defaults to 5.

scanner.nodeSelector

Specify a node selector label as label-key: label-value to force Scanner to only schedule on nodes with the specified label.

scanner.tolerations

If the node selector selects tainted nodes, use this parameter to specify a taint toleration key, value, and effect for Scanner.

scanner.dbNodeSelector

Specify a node selector label as label-key: label-value to force Scanner DB to only schedule on nodes with the specified label.

scanner.dbTolerations

If the node selector selects tainted nodes, use this parameter to specify a taint toleration key, value, and effect for Scanner DB.

scanner.resources.requests.memory

The memory request for the Scanner container. Use this parameter to override the default value.

scanner.resources.requests.cpu

The CPU request for the Scanner container. Use this parameter to override the default value.

scanner.resources.limits.memory

The memory limit for the Scanner container. Use this parameter to override the default value.

scanner.resources.limits.cpu

The CPU limit for the Scanner container. Use this parameter to override the default value.

scanner.dbResources.requests.memory

The memory request for the Scanner DB container. Use this parameter to override the default value.

scanner.dbResources.requests.cpu

The CPU request for the Scanner DB container. Use this parameter to override the default value.

scanner.dbResources.limits.memory

The memory limit for the Scanner DB container. Use this parameter to override the default value.

scanner.dbResources.limits.cpu

The CPU limit for the Scanner DB container. Use this parameter to override the default value.

Environment variables

You can specify environment variables for Sensor and Admission controller in the following format:

customize:
  envVars:
    ENV_VAR1: "value1"
    ENV_VAR2: "value2"

The customize setting allows you to specify custom Kubernetes metadata (labels and annotations) for all objects created by this Helm chart and additional pod labels, pod annotations, and container environment variables for workloads.

The configuration is hierarchical, in the sense that metadata defined at a more generic scope (for example, for all objects) can be overridden by metadata defined at a narrower scope (for example, only for the Sensor deployment).

Installing the secured-cluster-services Helm chart

After you configure the values-public.yaml and values-private.yaml files, install the secured-cluster-services Helm chart to deploy the per-cluster and per-node components (Sensor, Admission controller, and Collector).

To install Collector on systems that have Unified Extensible Firmware Interface (UEFI) and that have Secure Boot enabled, you must use eBPF probes because kernel modules are unsigned, and the UEFI firmware cannot load unsigned packages. Collector identifies Secure Boot status at the start and switches to eBPF probes if required.

Prerequisites
  • You must have generated RHACS init bundle for your cluster.

  • You must have the address and the port number that you are exposing the Central service on.

Procedure
  • Run the following command:

    $ helm install -n stackrox --create-namespace \
      stackrox-secured-cluster-services rhacs/secured-cluster-services \
      -f <name_of_cluster_init_bundle.yaml> \
      -f <path_to_values_public.yaml> -f <path_to_values_private.yaml> (1)
    1 Use the -f option to specify the paths for your YAML configuration files.

To deploy secured-cluster-services Helm chart by using a continuous integration (CI) system, pass the init bundle YAML file as an environment variable to the helm install command:

$ helm install ... -f <(echo "$INIT_BUNDLE_YAML_SECRET") (1)
1 If you are using base64 encoded variables, use the helm install …​ -f <(echo "$INIT_BUNDLE_YAML_SECRET" | base64 --decode) command instead.

Changing configuration options after deploying the secured-cluster-services Helm chart

You can make changes to any configuration options after you have deployed the secured-cluster-services Helm chart.

Procedure
  1. Update the values-public.yaml and values-private.yaml configuration files with new values.

  2. Run the helm upgrade command and specify the configuration files using the -f option:

    $ helm upgrade -n stackrox \
      stackrox-secured-cluster-services rhacs/secured-cluster-services \
      --reuse-values \ (1)
      -f <path_to_values_public.yaml> \
      -f <path_to_values_private.yaml>
    1 You must specify the --reuse-values parameter, otherwise the Helm upgrade command resets all previously configured settings.

    You can also specify configuration values using the --set or --set-file parameters. However, these options are not saved, and it requires you to manually specify all the options again whenever you make changes.

Installing RHACS on secured clusters by using the roxctl CLI

To install RHACS on secured clusters by using the CLI, perform the following steps:

  1. Install the roxctl CLI

  2. Install Sensor.

Installing the roxctl CLI

You must first download the binary. You can install roxctl on Linux, Windows, or macOS.

Installing the roxctl CLI on Linux

You can install the roxctl CLI binary on Linux by using the following procedure.

Procedure
  1. Download the latest version of the roxctl CLI:

    $ curl -O https://mirror.openshift.com/pub/rhacs/assets/3.74.9/bin/Linux/roxctl
  2. Make the roxctl binary executable:

    $ chmod +x roxctl
  3. Place the roxctl binary in a directory that is on your PATH:

    To check your PATH, execute the following command:

    $ echo $PATH
Verification
  • Verify the roxctl version you have installed:

    $ roxctl version

Installing the roxctl CLI on macOS

You can install the roxctl CLI binary on macOS by using the following procedure.

Procedure
  1. Download the latest version of the roxctl CLI:

    $ curl -O https://mirror.openshift.com/pub/rhacs/assets/3.74.9/bin/Darwin/roxctl
  2. Remove all extended attributes from the binary:

    $ xattr -c roxctl
  3. Make the roxctl binary executable:

    $ chmod +x roxctl
  4. Place the roxctl binary in a directory that is on your PATH:

    To check your PATH, execute the following command:

    $ echo $PATH
Verification
  • Verify the roxctl version you have installed:

    $ roxctl version

Installing the roxctl CLI on Windows

You can install the roxctl CLI binary on Windows by using the following procedure.

Procedure
  • Download the latest version of the roxctl CLI:

    $ curl -O https://mirror.openshift.com/pub/rhacs/assets/3.74.9/bin/Windows/roxctl.exe
Verification
  • Verify the roxctl version you have installed:

    $ roxctl version

Installing Sensor

To monitor a cluster, you must deploy Sensor. You must deploy Sensor into each cluster that you want to monitor. The following steps describe adding Sensor by using the RHACS portal.

Prerequisites
  • You must have already installed Central services, or you can access Central services on Red Hat Advanced Cluster Security Cloud Service (ACSCS).

Procedure
  1. On the RHACS portal, navigate to Platform ConfigurationClusters.

  2. Select + New Cluster.

  3. Specify a name for the cluster.

  4. Provide appropriate values for the fields based on where you are deploying the Sensor.

    • If you are deploying Sensor in the same cluster, accept the default values for all the fields.

    • If you are deploying into a different cluster, replace central.stackrox.svc:443 with a load balancer, node port, or other address, including the port number, that is accessible from the other cluster.

    • If you are using a non-gRPC capable load balancer, such as HAProxy, AWS Application Load Balancer (ALB), or AWS Elastic Load Balancing (ELB), use the WebSocket Secure (wss) protocol. To use wss:

      • Prefix the address with wss://.

      • Add the port number after the address, for example, wss://stackrox-central.example.com:443.

  5. Click Next to continue with the Sensor setup.

  6. Click Download YAML File and Keys to download the cluster bundle (zip archive).

    The cluster bundle zip archive includes unique configurations and keys for each cluster. Do not reuse the same files in another cluster.

  7. From a system that has access to the monitored cluster, unzip and run the sensor script from the cluster bundle:

    $ unzip -d sensor sensor-<cluster_name>.zip
    $ ./sensor/sensor.sh

    If you get a warning that you do not have the required permissions to deploy Sensor, follow the on-screen instructions, or contact your cluster administrator for assistance.

After Sensor is deployed, it contacts Central and provides cluster information.

Verification
  1. Return to the RHACS portal and check if the deployment is successful. If successful, when viewing your list of clusters in Platform ConfigurationClusters, the cluster status displays a green checkmark and a Healthy status. If you do not see a green checkmark, use the following command to check for problems:

    • On OpenShift Container Platform:

      $ oc get pod -n stackrox -w
    • On Kubernetes:

      $ kubectl get pod -n stackrox -w
  2. Click Finish to close the window.

After installation, Sensor starts reporting security information to RHACS and the RHACS portal dashboard begins showing deployments, images, and policy violations from the cluster on which you have installed the Sensor.