EKS - Amazon Elastic Kubernetes service onboarding

Mavvrik supports three ways to connect an EKS cluster, as shown in the Connect EKS Cluster screen:

Allocation-Based – recommended when you want quick onboarding and cost allocation visibility without deep cluster telemetry.
Appliance-Based – recommended when you need deeper telemetry and more detailed Kubernetes insights.
Auto-discover and connect — recommended when you want to connect clusters that are created frequently

Why this matters: these three methods trade off setup effort against depth of insight. Allocation-Based leans on AWS's own cost-allocation tags — fastest to set up, but limited to what AWS's billing data exposes. Appliance-Based deploys Mavvrik's own agent into the cluster via Helm, giving deeper per-pod telemetry at the cost of a more involved setup. Auto-discover automates the Allocation-Based approach for environments where clusters are created and torn down frequently, so nobody has to remember to onboard each new one manually.

Steps to onboard the EKS cluster(Appliance based)

Before onboarding your EKS account, you must have the admin rights on the cluster that you want to onboard and follow the steps below

Step 1: Enable Cost Allocation Tags

Access the AWS Portal, then search for and select Billing and Cost Management.

Navigate to Cost Allocation Tags and select AWS-generated cost allocation tags.

Search aws:eks:cluster-name & aws:autoscaling:groupName and enable the tag.

Enable EKS cluster and Auto Scaling cost allocation tags

Ref link: https://us-east-1.console.aws.amazon.com/costmanagement/home?region=us-east-1#/tags

Step-2: Install Kubectl and Helm.

Login to AWS console where EKS clusters are created
Click on Cloud Shell on top right corner
Select the cluster you wanna onboard by running the command below. Replace 'your-cluster-name' with the name of the specific cluster you want to onboard.
```
aws eks update-kubeconfig --region us-east-1 --name dx-enteryourclustername
```

If kubctl and helm are installed then skip the below steps and perform Step-3 (below).

Steps to install kubectl

#macOS
brew install kubectl

#Windows
winget install Kubernetes.kubectl

#Linux
curl -LO "https://dl.k8s.io/release/$(curl -L -s https://dl.k8s.io/release/stable.txt)/bin/linux/amd64/kubectl"
chmod +x kubectl
sudo mv kubectl /usr/local/bin/

Steps to install helm

brew install helm   # macOS
winget install Helm.Helm  # Windows

Step-3 : Steps to perform in Mavvrik

Go to the Menu
Click on Admin → Kubernetes

Navigate to Kubernetes administration in Mavvrik

Click on the EKS tab
Click on Connect > Appliance based
Enter your Prometheus URL

Steps to get Prometheus URL

Capture the Prometheus URL associated with the K8s cluster:

a. Enter the following URL in Mavvrik UI

http://<prometheus-server-name>.<namespace>:<prometheus-internal-port>

b. To find the Prometheus service, run the following command in Kubernetes shell

kubectl get svc -n prometheus

In URL Replace prometheus-server-name,namespace & prometheus-internal-port with your data
The following screenshot shows an example of a URL

Enter EKS cluster connection details in Mavvrik

Copy the mavvrik UI Step-4 Kubernetes Shell command and run it in aws console Cloud Shell

Run the Mavvrik onboarding script in AWS Cloud Shell

After executing the command the Kubernetes appliance will be installed.
Click on the Connect button in Mavvrik.

Instructions to Enable Resource Requests split cost allocation in Cost & Usage Reports (Allocation based)

Access the AWS Portal, then search for and select Billing and Cost Management.

Navigate to Cost Allocation Tags.

Click select all checkbox

Activate the selected cost allocation tags

Then click on Active button

To opt in to split cost allocation data
1. Open the Billing and Cost Management console at the main billing account level
2. In the navigation pane, choose Cost Management preferences in the left side panel.

c. Under General, in the Split cost allocation data section, choose between the following:

Amazon Elastic Kubernetes Service (Amazon EKS) to opt in to Amazon EKS only. For Amazon EKS, choose between the following:

Resource requests: This allocates your Amazon EC2 by Kubernetes pod CPU and memory resources only. This will encourage application teams to only provision what they need.

Include cost and usage data in your report
1. Open the Billing and Cost Management console at https://console.aws.amazon.com/costmanagement/.
2. In the navigation pane, under Legacy Pages, choose Cost and Usage Reports.
3. Edit an existing report created while onboarding AWS, in the Specify report details page, under Report content, select Split cost allocation data.

Include split cost allocation data in the Cost and Usage Report

Navigate to Mavvrik portal > Admin > K8s and connect the Allocation based cluster

Connect the allocation-based EKS cluster in Mavvrik

Steps to onboard the EKS cluster(Auto-discover and connect based)

Before onboarding your EKS account, you must have the admin rights on the cluster that you want to onboard and follow the steps below

Step 1: Enable Cost Allocation Tags

Access the AWS Portal, then search for and select Billing and Cost Management.

Navigate to Cost Allocation Tags and select AWS-generated cost allocation tags.

Search aws:eks:cluster-name & aws:autoscaling:groupName and enable the tag.

Ref link: https://us-east-1.console.aws.amazon.com/costmanagement/home?region=us-east-1#/tags

Step-2: Enable the discovery and connect cluster in Mavvrik

Login to Mavvrik Portal
Click on Menu
In Admin section, Click on Accounts
Click on Kubernetes

Tap on EKS
Click on the setting icon

then enable the EKS cluster

After enabling the cost allocation tags, wait for 24 hours for the cluster to be automatically onboarded

Troubleshooting

If the installation command fails, confirm that kubectl and Helm are connected to the intended EKS cluster.
If allocation data is unavailable, verify that the required AWS cost allocation tags and split cost allocation settings are enabled.