Skip to main content
Connect your Kubernetes clusters to Draftt to detect version compatibility issues, track configuration drift, and get actionable upgrade plans. Draftt uses a lightweight Helm-based collector deployed into your cluster.

​
Prerequisites

  • A Kubernetes cluster with kubectl configured and pointing to the correct context
  • Helm installed on your machine
  • A Draftt account with tenant admin permissions (to create an access key)

​
Setup

1

Select a cluster

In Draftt, go to Integrations > K8s Explorer > Add. Select the cluster you want to connect.
2

Add the Draftt Helm repo

Add the Draftt Helm chart repository and update:
helm repo add draftt-io https://draftt-io.github.io/helm-charts
helm repo update
3

Create a Draftt API access key

Navigate to Settings > Access Keys in Draftt. Create a new access key with tenant admin permissions. Save the key — you will need it in the next step.
4

Create a Kubernetes secret

Draftt Explorer is a collector tool. We recommend deploying it in a monitoring namespace (if one exists), but any namespace works.Create a secret that stores your Draftt API key securely:
kubectl create secret generic draftt-api-token \
  --from-literal=drafttApiToken=<YOUR_API_KEY> \
  -n monitoring
Make sure your kubectl context is pointing to the correct cluster before running this command.
5

Install the Helm chart

Install the Draftt Explorer chart:
helm install draftt-explorer draftt-io/draftt-explorer \
  --set appConfig.clusterIdentifier="<YOUR_K8S_CLUSTER_IDENTIFIER>" \
  --set appConfig.namespace="monitoring" \
  -n monitoring
Replace <YOUR_K8S_CLUSTER_IDENTIFIER> with a name that identifies this cluster in Draftt (e.g. production-us-east-1, staging-eu).
6

Validate the integration

Run the collector job manually to verify everything is working:
kubectl create job --from=cronjob/draftt-explorer draftt-test-run -n monitoring
Check the job logs to confirm data was sent to Draftt successfully. Once validated, the collector runs on its scheduled cron interval automatically.

​
What Draftt Collects

The Explorer is a read-only collector. It queries the Kubernetes API for resource metadata and sends it to Draftt for analysis. It does not modify any cluster resources. Draftt governs workloads (Deployments, StatefulSets, DaemonSets, Jobs, CronJobs), networking (Services, Ingress, NetworkPolicies), security (ServiceAccounts, Roles, ClusterRoles, RBAC bindings), configuration (ConfigMaps, Secrets), and storage (PersistentVolumes, PersistentVolumeClaims, StorageClasses).

​
Supported Environments

The Helm-based collector works with any Kubernetes distribution: managed services (EKS, AKS, GKE), self-managed clusters, and on-premises deployments. Connect multiple clusters by repeating the setup for each one.

​
Verifying Your Connection

After the test job completes, return to Integrations > K8s Explorer in Draftt. The connected cluster shows a status:
  • Healthy — The collector is running and Draftt is receiving data.
  • Unhealthy — Check that the API key secret exists in the correct namespace, the Helm chart is installed, and the CronJob is scheduled.

​
Troubleshooting

Job fails with authentication error: Verify the draftt-api-token secret exists in the correct namespace and contains a valid API key with tenant admin permissions. No data appearing in Draftt: Confirm the clusterIdentifier value matches what you expect in the Draftt dashboard. Check the job logs with kubectl logs job/draftt-test-run -n monitoring. Helm install fails: Ensure the Draftt Helm repo was added and updated (helm repo update). Verify the target namespace exists (kubectl create namespace monitoring if needed).