Skip to content
Kedify Docs

Install with Helm

Install the Kedify Agent, KEDA and HTTP Scaler in one Helm release. For an existing upstream KEDA installation, start with migration.

Prerequisites

Get the command from the Dashboard

  1. In Clusters, select Connect cluster. For an existing cluster, open its summary and select Installation.
  2. Select Helm at the top right. Choose components, enter a cluster name and select an Agent API key.
  3. For isolated operators, select the multi-tenant KEDA topology and configure tenants. For remote workloads, configure multi-cluster mode. Review the generated steps for every selected installation.
  4. Add the Helm repository, then run the installation steps in order in the intended Kubernetes context.

View credentials opens the credential section, where you can select an API key, manage organization keys, copy the organization ID or API key, and preview or copy Secret YAML. Helm creates keda/kedify-agent automatically from the selected key; you do not need to apply a separate Secret for the generated Helm command. Treat the copied command as sensitive because it includes the API key.

For a Git-managed installation, switch to Argo CD or Flux. Those exports reference credentials supplied separately.

Install Kedify dialog with the Helm method selected, component selection and cluster configuration

Prepare Installation

Follow these steps to install Kedify using Helm manually.

Getting Organization ID

In the Kedify dashboard, navigate to Organization and select Details. Copy the ID.

Alternatively, if you already have an existing installation, you can read the Organization ID using:

Terminal window
kubectl set env -nkeda deployment/kedify-agent --list | grep KEDIFY_ORGANIZATION_ID

Getting API Key

If you already have a Kedify Agent deployed, you can retrieve your pre-provisioned API key from a Secret in the same Kubernetes cluster:

Terminal window
kubectl get secret -n keda kedify-agent -o=jsonpath='{.data.apikey}' | base64 --decode

If you don't have an existing Agent, you can generate a new API Key in the Kedify dashboard. Navigate to Organization and select API Keys. Click Create Agent Key and copy the key.

Install

These commands explicitly enable KEDA and HTTP Scaler in the Agent chart. All charts are available in the Kedify charts repository.

Kedify Helm Repository

To install Kedify using Helm, you first need to add the Kedify Helm repository to your local Helm configuration.

Terminal window
helm repo add kedifykeda https://kedify.github.io/charts
helm repo update kedifykeda

Quick Installation

Replace clusterName, $YOUR_ORG_ID and $YOUR_API_KEY with your installation values.

This command installs the Kedify Agent along with KEDA and the Kedify HTTP Scaler. If you don't want to install KEDA or the HTTP Scaler, you can disable them by setting keda.enabled=false or keda-add-ons-http.enabled=false.

You can also install additional components such as the OTel Scaler and Kedify Predictor. You can enable them by setting otel-add-on.enabled=true or kedify-predictor.enabled=true respectively.

Terminal window
helm upgrade --install kedify-agent kedifykeda/kedify-agent \
--namespace keda \
--create-namespace \
--version v0.8.3 \
--set clusterName="my-cluster" \
--set agent.orgId="$YOUR_ORG_ID" \
--set agent.apiKey="$YOUR_API_KEY" \
--set keda.enabled=true \
--set keda-add-ons-http.enabled=true \
--set otel-add-on.enabled=false \
--set kedify-predictor.enabled=false

Verify the installation

Terminal window
kubectl get pods -n keda
kubectl get apiservice v1beta1.external.metrics.k8s.io

Confirm the enabled components are ready and the external metrics API reports AVAILABLE=True. Check the cluster connection in the Dashboard, then configure a scaler and verify workload behavior. Autoscaling Checks can exercise the signal-to-replica path.

Other installation paths

Helm values · Capability enablement · KedifyConfiguration · Access and data boundaries