zalando-kubectl 1.31.1.237

Kubectl wrapper in Python with OAuth token auth

Kubernetes CLI (kubectl) wrapper in Python with OAuth token authentication.

This wrapper script zkubectl serves as a drop-in replacement for the kubectl binary:

• it downloads the current kubectl binary from Google

• it generates a new ~/.kube/config with an OAuth Bearer token acquired via zign.

• it passes through commands to the kubectl binary

Installation

Requires Python 3.6+.

$ sudo pip3 install --upgrade zalando-kubectl

Usage

You can directly login to a known Kubernetes API server endpoint:

$ zkubectl login https://my-api-server.example.org
$ zkubectl cluster-info

You can also configure a Cluster Registry to look up clusters by ID:

$ zkubectl configure --cluster-registry=https://cluster-registry.example.org
$ zkubectl login my-cluster-id

The Cluster Registry needs to provide the following HTTP API for this to work:

$ curl -H "Authorization: Bearer $(zign tok)" https://cluster-registry.example.org/kubernetes-clusters/my-cluster-id
{
    "api_server_url": "https://my-api-server.example.org"
}

There is an additional convenience command to open the kube-web-view dashboard in the browser:

$ zkubectl dashboard
Opening https://kube-web-view.zalando.net/ ..

Unit Tests

Run unit tests with Tox:

$ sudo pip3 install tox
$ tox

Local Changes

It’s recommended to have a virtualenv for the project. The project uses Black for code formatting, please configure your editor to use it.

Go to the project dir and install dependencies into virtual test environment

$ cd <project-path>
# create new virtual environment if not yet
$ python -m venv test_environment
# enter virtual environment
$ source ./test_environment/bin/activate
# check pip is executed fron this virtual environment
(test_environment) $ which pip
<project-path>/test_environment/bin/pip

(test_environment) $ pip install -r requirements.txt
(test_environment) $ pip install --editable .

Now the code change will just be reflected in the zkubectl binary

(test_environment) $ zkubectl <whatever>

As an alternative for creating an environment you can run local changes from zalando_kubectl

$ cd <project-path>
$ python -m zalando_kubectl <whatever>