Codefresh for CD

Continuous deployment (CD) with Codefresh pipelines

Codefresh is a Continuous Integration/Delivery (CI/CD) solution. This article reviews main concepts around CD, and how Codefresh supports and implements them.

For a review of CI concepts, see Codefresh for CI.

Connecting to Kubernetes

Connecting your Kubernetes cluster to the CI/CD platform is the first step in continuous deployment. Codefresh integrates with any known cluster provider for Kubernetes through a few simple steps. Connect your Google, Azure, or Amazon Kubernetes cluster to Codefresh through simple integration steps. If your Kubernetes cluster is not in our list of cluster providers, you can manually enter your cluster settings and add it as a generic Kubernetes cluster.

See Connecting a Kubernetes cluster.

Deploying to Kubernetes

Codefresh offers a variety of options for you to choose from when deploying to Kubernetes. Deploy to Kubernetes from the Codefresh UI, or programmatically through dedicated steps in pipelines, avoiding the need for kubectl commands.

On-demand deployment

For quick and easy deployment, you can deploy on-demand from the Codefresh UI.

Dedicated steps in pipelines

For more flexibility, we have dedicated steps for pipelines: the deploy step, and the more advanced cf-deploy-kubernetes step which enables simple templating on Kubernetes manifests.

Codefresh pipelines also support Kustomize and Helm for deployments through freestyle steps.

Finally, if you are familiar with and still want to work with kubectl, run your own custom kubectl commands in a freestyle step. Read more in kubectl.

See Deployment options for Kubernetes.


kubectl is the command line interface for managing Kubernetes clusters. Running custom kubectl commands in a freestyle step gives maximum flexibility with cluster deployments.

Codefresh automatically sets up your config context with your connected clusters. The config context is the value of the $CF_KUBECONFIG_PATH variable, which expands to /codefresh/volume/sensitive/.kube/config within the shared step volume.

Codefresh has a public Docker image for kubectl at Docker Hub that you can use.

Because Codefresh automatically sets up your kubeconfig files with the information from your cluster integrations, you can modify the current config context and run any kubectl command you want applied to that context. For example, leverage the parallel capability of Codefresh pipelines to create two Docker images and deploy them to two different clusters with custom kubectl commands.

See Running custom kubectl commands.

Helm deployments

Codefresh supplies a built-in Helm repository with every Codefresh account. And supports, besides public HTTP repositories, several private, authenticated Helm repositories.

Connect to Helm repositories
In addition to the official Helm repositories which are displayed in the Helm charts, Codefresh allows you to connect with any external Helm repository through simple integrations. You can then inject the Helm repository context into your pipelines by selecting the repository name.

Build Helm charts
Install Helm charts from Helm repositories, or build a new one.

As with steps for Kubernetes deployments, Codefresh has a dedicated step for Helm deployments. The Helm step easily integrates Helm in Codefresh pipelines, and can authenticate, configure, and execute Helm commands.

The Helm step can operate in one of three modes covering most scenarios:

  • Install the chart into a Kubernetes cluster. This is the default mode if not explicitly set.
  • Package the chart and push it to the defined repository.
  • Set up authentication and add the repo to Helm. This is useful if you want to write your own Helm commands using the freestyle step’s commands property, but you still want the step to handle authentication.

Deploy Helm charts
Deploy the Helm chart to a Kubernetes cluster, Helm repo, or both.

See Using Helm in Codefresh pipelines, Using managed Helm repos, and Helm charts and repositories.


Dashboards are key to providing the right information at the right time. Codefresh makes it easy to both access and visualize critical information for any resource, at any stage and level, and for anyone, from managers to DevOps engineers.

Our operational dashboards expose the most critical application and environmental information to developers for troubleshooting without needing assistance from the DevOps teams. Our analytics dashboards expose key statistics and metrics around builds and deployments for product owners and management alike.

You can customize all dashboards to display just the information most relevant to your business issues.


The Kubernetes (Kubernetes Services) dashboard displays the clusters in your environments, their state, and actions if you have the required access privileges.

This dashboard is a centralized location from which to view, add, modify, and remove Kubernetes services. You can deploy services with images from the Codefresh registry or from external Docker registries.
See Accessing the Kubernetes dashboard.

Helm dashboards

For Helm, we have a Helm Board, and a Helm Releases dashboard.

  • Helm Boards The Helm Board is a special environment board that tracks applications as they move within your infrastructure. For example, between Dev, QA and Prod environments.
    Here you can see the lifecycle of the application, and at the same time, use it as a tool to shift Helm releases between environments.

    You can create as many Helm Boards as you want, and customize the Helm Boards with the columns to display, each column being a Helm-enabled Kubernetes cluster.
    See Using the Helm Board.

  • Helm Releases The Helm Releases dashboard provides a unique view into your production Kubernetes cluster, and options to manage deployed releases. See the current status of your cluster, including the currently deployed releases, their previous revisions with change tracking. If needed, you can even roll back to a previous release.
    See View Helm releases and release information.

Environment dashboard

The Kubernetes dashboard shows cluster status and focuses more on the service aspects such as pods and Docker images. The Helm Board offers an application-level view of your cluster, as it applies to Helm deployments.

To bridge this gap, Codefresh also offers an Environment dashboard that combines information for both Kubernetes and Helm releases. You can see cluster status in the context of the pipeline builds that run on and affect the cluster. The Environment dashboard links between the status of the cluster and the status of the builds.

See Environment dashboard.

Home Dashboard

The Home Dashboard presents system-wide highlights in real-time for GitOps deployments, Argo CD Workflows, and pipelines, making it an ideal tool for management.

  • GitOps Dashboard: Get insights into important KPIs for entities across runtimes and clusters, in the same location. View status of runtimes and managed clusters, deployment history, and most active applications.
  • Argo Workflows Dashboard: Get insights into Argo Workflows and Delivery Pipelines.
  • Pipelines Dashboard: Get insights into pipeline activity.

See Home Dashboard.

DORA metrics

The DORA (DevOps Research and Assessment) metrics dashboard pulls out critical metrics for our GitOps deployments. The metrics include deployment frequency, lead time for changes, mean time to recovery, and change failure rate. The metric graphs show performance by default for the last 90 days, with the option to select daily/weekly/monthly granularity views.

See DORA metrics.

Codefresh for CI
Codefresh for GitOps
Concepts in Codefresh