Skip to main content

Local Script Installation

How It Works​

install.sh is an interactive bash script that performs the following steps in order:

  1. Detects and installs Helm (if not already installed locally)
  2. Uses Helm to install Tetragon to the cluster (if not already deployed)
  3. Deploys K8s Sentinel to the sentinel-system namespace via kubectl apply
  4. Uses Helm to install Trivy Operator into sentinel-system (v0.61+, if not already deployed in the cluster)

The entire installation runs locally, printing progress messages at each step, making it easy to observe the installation status. Ideal for quick evaluation or development/test environments.


Step 1: Clone the Repository​

git clone https://github.com/cooloo9871/K8s_Sentinel.git
cd K8s_Sentinel/deploy

After cloning, navigate to the deploy/ directory where the install.sh script is located.


Step 2: Grant Execute Permission​

chmod +x install.sh

Why: Linux systems do not allow direct execution of scripts downloaded from the network by default. You must explicitly grant the execute bit with chmod +x before running with ./install.sh. This is a security mechanism to prevent accidental execution of unknown scripts.


Step 3: Run the Installation Script​

./install.sh

The script executes the following flow:

  1. Prepare Helm - When Helm is needed to install Tetragon or Trivy Operator and helm is not found locally, the script downloads the latest stable release into a temp directory for its own use (nothing is installed system-wide; cleaned up when the script exits)
  2. Deploy Tetragon - Skipped when a Tetragon DaemonSet already exists in any namespace (not only kube-system since v0.69); otherwise installed from the official Cilium Helm Chart as a Helm release (helm install since v0.64.1, so a later helm upgrade works) with hostNetwork=false and tetragon.grpc.address=0.0.0.0:54321. Whether or not Tetragon was installed just now, the script applies tetragon-grpc-policy.yaml so that only K8s Sentinel can reach the gRPC port (see Installing Tetragon for the reason)
  3. kubectl apply the K8s Sentinel manifest - Creates the required Namespace, ServiceAccount, ClusterRole, ClusterRoleBinding, Deployment, and Service resources
  4. Deploy Trivy Operator - Skipped when a Trivy Operator Deployment already exists in any namespace of the cluster; otherwise installed into sentinel-system with helm install, using the trivy-operator-values.yaml in the same directory (see Installing Trivy Operator)

Step 4: Monitor Installation Output​

During installation, the script prints progress messages for each phase:

Output PhaseDescription
[helm] not found, downloading to temp dir...Helm is not installed locally, so it is downloaded to a temp directory (only appears when needed)
[tetragon] already installed, skippingTetragon already exists in the cluster, skipped
[tetragon] not found, installing latest version...Deploying Tetragon and waiting for the DaemonSet to become ready; prints [tetragon] ready when done
[sentinel] applying manifests...Applying K8s Sentinel Kubernetes resources and waiting for the Deployment to become ready
[trivy-operator] already installed, skippingTrivy Operator already exists in the cluster, skipped
[trivy-operator] not found, installing latest version...Installing Trivy Operator and waiting for the Deployment to become ready; prints [trivy-operator] ready when done
Done. Access via:All steps completed

After installation, the script displays a suggested port-forward command:

kubectl port-forward -n sentinel-system svc/sentinel 8080:80

Copy and run this command locally to open the K8s Sentinel management UI in your browser.


Step 5: Verify Deployment Status​

kubectl get all -n sentinel-system

Expected result:

  • Pod status is Running
  • Service type is ClusterIP with an internal IP assigned
NAME READY STATUS RESTARTS AGE
pod/sentinel-7d9f8b6c4-xxxxx 1/1 Running 0 2m

NAME TYPE CLUSTER-IP EXTERNAL-IP PORT(S) AGE
service/sentinel ClusterIP 10.96.123.45 <none> 80/TCP 2m

NAME READY UP-TO-DATE AVAILABLE AGE
deployment.apps/sentinel 1/1 1 1 2m

If the Pod does not reach Running state, check for detailed error messages:

kubectl describe pod -n sentinel-system
kubectl logs -n sentinel-system deployment/sentinel

Production Environment Note

The script automatically downloads Helm and modifies cluster configuration (adding Tetragon and Trivy Operator, creating RBAC resources, etc.). In production environments, such automated operations may bypass change review processes. Consider using Kubernetes Job Installation instead for better operational transparency and control.