Learn how to install the StormForge Agent on a cluster


You’ll need:

  • A StormForge Account
  • Kubernetes version 1.16 or later
  • Helm version 3.8 or later
  • kubectl properly configured for your cluster
  • Cluster-Admin access (for installation only)
  • A Kubernetes cluster and the cluster name to install the Agent into

In your organization’s allowlist, allow access to StormForge services endpoints on port 443.

  • Grant Kubernetes clusters access to these endpoints, which are the minimum required for installing the StormForge Agent:

      (*If you configure the Agent to use an internal registry, you don’t need to grant cluster access to this endpoint).
  • Grant workstations access these endpoints:


Prepare to install

Install the StormForge CLI tool

The StormForge CLI is used to authenticate, retrieve credentials, and register the StormForge Agent on your Kubernetes cluster.

Download and install the CLI tool:

Linux Binary Download:

Amd64 | Arm64

Linux install command:

# Automatically selects either AMD64 or ARM64 architecture, downloads
# the appropriate binary, then moves it to a location in PATH
{ [ "$(uname -sm)" = "Linux x86_64"  ] && curl -L | tar -xz; } ||
{ [ "$(uname -sm)" = "Linux aarch64" ] && curl -L | tar -xz; } &&
sudo mv stormforge /usr/local/bin/

Windows Binary Download:

Amd64 | Arm64

PowerShell install command:

[Net.ServicePointManager]::SecurityProtocol = [Net.SecurityProtocolType]::Tls12
If ($env:PROCESSOR_ARCHITECTURE -eq "AMD64") { Invoke-WebRequest -Uri "" -Outfile "" }
If ($env:PROCESSOR_ARCHITECTURE -eq "ARM64") { Invoke-WebRequest -Uri "" -Outfile "" }
Expand-Archive "" "." -WarningVariable $w; if ($w.Count -eq 0) { Remove-Item "" } 

Then, move stormforge.exe to a folder defined in your Windows PATH environment variable.

macOS Binary Download:

Amd64 | Arm64

macOS install command:

# Automatically selects either AMD64 or ARM64 architecture, downloads
# the appropriate binary, then moves it to a location in PATH
{ [ "$(uname -sm)" = "Darwin x86_64" ] && curl -L | tar -xz; } ||
{ [ "$(uname -sm)" = "Darwin arm64"  ] && curl -L | tar -xz; } &&
sudo mv stormforge /usr/local/bin/


brew install thestormforge/tap/stormforge

We provide a container image at (for linux/amd64 and linux/arm64):

docker pull

We recommend setting the STORMFORGE_TOKEN environment variable (from auth create -o token) if you want to use the container image directly.

Log in

From the command line, log into your StormForge account:

stormforge login

Alternatively, you can get a one-time use code to enter into a browser from another device:

stormforge login --url

At any time, you can check to see that you can communicate with the API:

stormforge ping

After you log in, go back to your command line session.

Generate an access credential

If the Agent is already installed on a cluster, you can reuse the existing credential: locate the file that contains the credential, and skip to Install the StormForge Agent later in this topic.

If the Agent doesn’t exist on your estate, run the following command.

  • Replace CREDENTIAL_NAME with a name that will help you identify the credential (such as the name of the cluster you’re installing on).
  • Replace CREDENTIAL_FILE with a filename that will help you to identify the file that contans the credential (no filename extension is needed).
stormforge auth create CREDENTIAL_NAME > CREDENTIAL_FILE

The credentials file should look something like this:

   clientID: CLIENT_ID
   clientSecret: CLIENT_SECRET

Install the StormForge Agent

Note: Some installations require additional steps. If any of the following items are true, go to Install the StormForge Agent - advanced scenarios:

  • You installed an earlier 2.0.x version of the stormforge-agent Helm chart
  • You use a proxy server or a private container registry
  • You cannot run helm install

To install the Agent, run the following command:

  • Replace CREDENTIAL_FILE with the name of the file that contains your access credential.
  • Replace CLUSTER_NAME with the name of the cluster you’re installing the Agent on.
  • Include any extra arguments from the Additional agent installation options table below.
helm install stormforge-agent oci:// \
  --namespace stormforge-system \
  --create-namespace \
  --values CREDENTIAL_FILE \
  --set clusterName=CLUSTER_NAME

Additional Agent installation options

Option Description
--set manageAuthSecret=false Set to false to manage the StormForge Agent secret with a third-party credential manager. For details and the Agent secret format, see this guide.
--set openshift=true Install on Red Hat OpenShift Container Platform version 4.x and later. See this guide for details.
--set 'workload.allowNamespaces={namespace1,namespace2}' Set an allow list to collect metrics only on workloads in specific namespaces. Overrides a denyNamespaces list. Regex permitted, such as .*-system.
--set 'workload.denyNamespaces={namespace1,namespace2}' Set a deny list so that no metrics are collected on workloads in these namespaces. Ignored if allowNamespaces is provided. Regex permitted, such as .*-system.

The Agent is now sending metrics to the cloud service. In approximately one hour, a preliminary recommendation for each workload will be available for you to view in the Optimize Live UI.

Validate the installation

To see your cluster and workloads:

  1. Log into with your StormForge account.
  2. In the left navigation, click Workloads.
  3. Optional: Refine your search by cluster, namespace, or workload name; press Enter, and review the results.

If you don’t see what you expect, try troubleshooting the Agent.

Should you install the StormForge Applier?

Install the StormForge Applier if you plan to:

  • Deploy recommendations on demand. You can apply a single recommendation in any environment as you experiment with recommendations or if you need to quickly deploy a recommendation outside of a schedule.
  • Deploy recommendations automatically. Define a schedule to receive recommendations, and Optimize Live will generate and deploy them. This option enables you to immediately realize the benefits of right-sizing your workloads without manual intervention.

You do not need to install the StormForge Applier if you plan to:

  • Deploy recommendations as part of a GitOps workflow (for example, in a staging or production environment)
  • Export the patches manually from the Optimize Live web application.

Install the Applier

First, make sure that the Agent is installed:

kubectl get pods -A | grep 'stormforge-agent'

Next, install the Applier:

  • If you cannot run helm install, go to this section in the Install the StormForge Agent - advanced scenarios topic.
helm install stormforge-applier \
  oci:// \
  -n stormforge-system
Additional Applier installation options
Option Description
--values applier-rbac.yaml To apply patches to a KEDA-owned HPA, you must grant additional RBAC permission to the Applier. See this guide for details.

Then, to validate the installation, run:

kubectl get pods -A | grep 'stormforge-applier'

The output should look something like this:

stormforge-system    stormforge-applier-8554758f5b-whbc4             1/1     Running   0              20s

Additional Agent installation considerations

Workload discovery and metrics collection

In a default Agent installation:

  • The Agent discovers and gathers metrics for all DaemonSet, HPA, Pod, ReplicaSet, ReplicationController, and StatefulSet types across all the namespaces on the cluster on which it’s installed.
  • Custom resources are ignored and no metrics are collected on them. If you have custom resources in your environment or if you prefer not to have workloads discovered automatically, contact for guidance on how to modify the Helm chart.
Role-based access control

By default:

  • The Agent applies the Kubernetes view role and creates a ClusterRoleBinding, granting read-only permissions to all resources in the cluster.
  • The optional Applier component uses the edit role, enabling it to patch all optimizable workloads.

What happens after you install?

First 7 days: metrics collection and preliminary recommendations

It typically takes about 7 days to collect enough metrics to generate an optimal recommendation (referred to as a complete recommendation) for you to apply.

Metrics collection on workloads begins as soon as the Agent is installed.

For the first 7 days after installation, Optimize Live generates preliminary recommendations to give you an evolving preview of potential optimization impact. They can be viewed, but should not be applied because they’re not based on a complete set of metrics. They’re generated:

  • Hourly, for the first 24 hours after installing the Agent
  • Once daily, on days 2 to 7 after installing the Agent

After 7 days: viewing and applying recommendations

After we collect 7 days of metrics, you’ll receive a complete recommendation once daily (the default schedule).

You can change this schedule at any time: On a workload’s details page, click Config, and set the Recommendation Schedule.

  • View and configure recommendations: You can view the recommendations on the Workloads page or on a workload’s details page.

  • Apply recommendations: By default, recommendations are not applied automatically, giving you control over how they’re applied.

    When you are satisfied that the recommendations are safe to apply, you can enable automatic deployment on the Config page for some or all of your workloads.

Helm charts and troubleshooting

To learn more about the Agent and Applier Helm charts and steps to troubleshoot, you can run the following commands:

helm show readme oci://
helm show readme oci://

To learn about other parameters that you can configure, run the following commands:

helm show values oci://
helm show values oci://
Last modified November 13, 2023