This page is outdated. Please check for up-to-date documentation.


Banzai Cloud's CI/CD solution runs entirely on top of Kubernetes, natively (by talking directly with the Kubernetes API). Its build steps are executed in a Kubernetes cluster that the first build step provisions. All builds are executed in ephemeral, isolated Namespaces, which are dynamically cleaned up and provisioned. Each build step is executed as a Kubernetes Pod in its namespace. Between build steps, data is persisted via a Persistent Volume. This is called a workspace.

To specify which cluster (context) you'd like to run your steps on, you must specify a cluster key in your pipeline.yaml. In its most basic form, a cluster key specifies only the name of an existing cluster to use, but it can also create a cluster if one did not previously exist. The following code snippet is an example configuration for an existing cluster:

  name: my-cluster

To create clusters, check out our guide to the Pipeline plugin.