Installing from AWS Marketplace

This guide provides step-by-step instructions to deploy Kubesense on an Amazon EKS (Elastic Kubernetes Service) cluster using kubectl, eksctl, and helm.

Prerequisites

Before deploying Kubesense, ensure the following:

You have an Amazon EKS cluster running.
You have identified the target cluster where Kubesense will be deployed (this is where sensors will be installed).
You have access to AWS CLI, eksctl, and kubectl installed on your local machine.
You have Helm installed on your local machine.

Server Deployment

Step 1: Create the `kubesense` Namespace

First, create a namespace in your Kubernetes cluster where the Kubesense resources will be deployed.

kubectl create namespace kubesense

Step 2: Create IAM Service Account for Kubesense

Use the following command to create an IAM service account for the kubesense namespace and attach the necessary AWS policies for marketplace metering and license management.

Replace <ENTER_YOUR_CLUSTER_NAME_HERE> with the name of your EKS cluster.

eksctl create iamserviceaccount \
    --name kubesense \
    --namespace kubesense \
    --cluster <ENTER_YOUR_CLUSTER_NAME_HERE> \
    --attach-policy-arn arn:aws:iam::aws:policy/AWSMarketplaceMeteringFullAccess \
    --attach-policy-arn arn:aws:iam::aws:policy/AWSMarketplaceMeteringRegisterUsage \
    --attach-policy-arn arn:aws:iam::aws:policy/service-role/AWSLicenseManagerConsumptionPolicy \
    --approve \
    --override-existing-serviceaccounts

This command will create the necessary IAM roles and policies required for the Kubesense AI platform to operate.

Step 3: Authenticate with AWS ECR

Next, authenticate with Amazon ECR (Elastic Container Registry) to pull the Kubesense Helm chart from the private repository.

aws ecr get-login-password \
    --region us-east-1 | helm registry login \
    --username AWS \
    --password-stdin 709825985650.dkr.ecr.us-east-1.amazonaws.com

Step 4: Pull the Kubesense Helm Chart

Create a directory to store the Helm chart and download the chart from the specified ECR repository.

mkdir kubesense && cd kubesense
helm pull oci://709825985650.dkr.ecr.us-east-1.amazonaws.com/kubesense-ai/kubesense --version 1.0.20

Step 5: Extract the Helm Chart

Extract the Helm chart by running the following commands:

tar xf $(pwd)/* && find $(pwd) -maxdepth 1 -type f -delete

This will extract the Helm chart into the current directory and remove any unnecessary files.

Step 6: Install Kubesense

Create kubesense-values.yaml file with below values.

NOTE: The default configuration for deploying Kubesense server components is optimized for environments with 8 vCPU and 16 GiB of memory. If your environment has resource constraints, adjust the resource allocation accordingly to match your available system capacity.

global:
  deploymentType: "server"
  cluster_name: <your_aws_cluster_name>
  dashboardHostName: <your_dashboard_host>
  aws_account_id: <your_aws_account_id>
  aws_region: <your_aws_region>

aggregator:
  resources:
    limits:
      cpu: 750m
      memory: 500Mi
    requests:
      cpu: 500m
      memory: 200Mi

clickhouse:
  resources:
    limits:
      cpu: 2000m
      memory: 4000Mi
    requests:
      cpu: 1000m
      memory: 3000Mi

kubecol:
  enabled: true
  resources:
    limits:
      cpu: 750m
      memory: 2000Mi
    requests:
      cpu: 500m
      memory: 1500Mi

mysql:
  resources:
    limits:
      cpu: 1000m
      memory: 1536Mi
    requests:
      cpu: 750m
      memory: 1024Mi

victoria-metrics-single:
  server:
    retentionPeriod: 7d
    resources:
      limits:
        cpu: 500m
        memory: 750Mi
      requests:
        cpu: 400m
        memory: 500Mi

victoria-metrics-agent:
  resources:
    limits:
      cpu: 500m
      memory: 750Mi
    requests:
      cpu: 400m
      memory: 500Mi

# Sensor Components
otel-agent:
  resources:
    limits:
      cpu: 500m
      memory: 768Mi
    requests:
      cpu: 250m
      memory: 512Mi

logsensor:
  resources:
    limits:
      cpu: 200m
      memory: 500Mi
    requests:
      cpu: 100m
      memory: 250Mi

resources:
  limits:
    cpu: 300m
    memory: 300Mi
  requests:
    cpu: 100m
    memory: 100Mi

Parameters:

cluster_name: The name of your Kubernetes cluster.
dashboardHostName: The hostname for the Kubernetes dashboard. Example: kubesense.your_domain.com
aws_account_id: Your AWS account ID where the infrastructure resources (such as EKS, S3, etc.) are managed.
aws_region: The AWS region where kubesense resources are deployed, e.g., us-west-2, us-east-1, etc.

Install the Kubesense Helm chart in the kubesense namespace using the Helm command.

helm install kubesense --namespace kubesense -f ./kubesense-values.yaml ./kubesense

This will deploy Kubesense in your EKS cluster under the kubesense namespace.

Step 7: Verify the Deployment

Once the installation is complete, you can verify that the resources are deployed correctly by checking the status of the pods in the kubesense namespace:

kubectl get pods -n kubesense

You should see the Kubesense pods running in your cluster.

Sensor Deployment

Identify the cluster you want to monitor. Install kubesense sensors into the cluster.

Create a directory to store the Helm chart and download the chart from the specified ECR repository.

mkdir sensor && cd sensor
helm pull oci://709825985650.dkr.ecr.us-east-1.amazonaws.com/kubesense-ai/kubesensor --version 1.0.7

Extract the Helm Chart

Extract the Helm chart:

tar xf $(pwd)/* && find $(pwd) -maxdepth 1 -type f -delete

Install sensors

Install the Kubesense Helm chart in the kubesense namespace using the Helm command.

helm install kubesensor --namespace kubesense -f ../kubesense-values.yaml ./kubesensor

After successfully deploying Kubesense on your Amazon EKS cluster, follow these steps to access and explore the platform. This section will guide you through logging in, accessing the dashboard, and navigating the core features to monitor resources deployed in the cluster.

Access the Kubesense Web Interface:

Open your web browser and navigate to the Kubesense platform using the provided URL. This could be either a public DNS endpoint or a load balancer URL that you set up during the deployment of Kubesense on your EKS cluster.

Example URL: https://<your-kubesense-dashboard-url>
Login with Default Credentials:

Use the default login credentials to sign in:
- Email Address: admin@kubesense.ai
- Password: tyke@AI123$
Once you enter the credentials, click Login.
Change the Default Password (Recommended):

Upon logging in for the first time, it is highly recommended to change the default password for security purposes.
- Hover on the User Avatar (At the left-bottom of sidebar), Open Profile screen.
- Choose Change Password, enter your current password (tyke@AI123$), and set a new, secure password.
- Click Save Changes to confirm the change.
Explore the Dashboard:

After logging in, you will be directed to the main dashboard. Here you can view traces of your application, Logs, and track Infrastructure performance of applications. Follow documentation to understand various capabilities of Kubesense.
Logging Out:

Once you are done exploring, you can safely log out by clicking the Logout button, found in the left-bottom corner of the dashboard under your profile icon.

Uninstalling kubesense

To remove all resources associated with Kubesense, delete the entire namespace with the following command:

kubectl delete namespaces kubesense

Prerequisites​

Server Deployment​

Step 1: Create the kubesense Namespace​

Step 2: Create IAM Service Account for Kubesense​

Step 3: Authenticate with AWS ECR​

Step 4: Pull the Kubesense Helm Chart​

Step 5: Extract the Helm Chart​

Step 6: Install Kubesense​

Parameters:​

Step 7: Verify the Deployment​

Sensor Deployment​

Extract the Helm Chart​

Install sensors​

Login and Explore​

Uninstalling kubesense​