Kubernetes Documentation

Note

This Kubernetes driver will be subject to change from community feedback. How to map the core assets (pods, clusters) to API entities will be subject to testing and further community feedback.

Kubernetes is an open source orchestration system for Docker containers. It handles scheduling onto nodes in a compute cluster and actively manages workloads to ensure that their state matches the users declared intentions. Using the concepts of “labels” and “pods”, it groups the containers which make up an application into logical units for easy management and discovery.

../../_images/kubernetes.png

Authentication

Authentication currently supported with the following methods:

Instantiating the driver

from libcloud.container.types import Provider
from libcloud.container.providers import get_driver

cls = get_driver(Provider.KUBERNETES)

# 1. Client side cert auth
conn = cls(
    host="192.168.99.103",
    port=8443,
    secure=True,
    key_file="/home/user/.minikube/client.key",
    cert_file="/home/user/.minikube/client.crt",
    ca_cert="/home/user/.minikube/ca.crt",
)

# 2. Bearer bootstrap token auth
conn = cls(key="my_token", host="126.32.21.4", ex_token_bearer_auth=True)

# 3. Basic auth
conn = cls(key="my_username", secret="THIS_IS)+_MY_SECRET_KEY+I6TVkv68o4H", host="126.32.21.4")

for container in conn.list_containers():
    print(container.name)

for cluster in conn.list_clusters():
    print(cluster.name)

Instantiating the driver (minikube installation - cert file auth)

This example shows how to connect to a local minikube Kubernetes cluster which utilizes certifcate based authentication.

import libcloud.security
from libcloud.container.types import Provider
from libcloud.container.providers import get_driver

# Disable cert vertification when running minikube locally using self signed
# cert
libcloud.security.VERIFY_SSL_CERT = False

cls = get_driver(Provider.KUBERNETES)

# You can retrieve cluster ip by running "minikube ip" command
conn = cls(
    host="192.168.99.103",
    port=8443,
    secure=True,
    key_file="/home/user/.minikube/client.key",
    cert_file="/home/user/.minikube/client.crt",
    ca_cert="/home/user/.minikube/ca.crt",
)

for container in conn.list_containers():
    print(container.name)

for cluster in conn.list_clusters():
    print(cluster.name)

Instantiating the driver (minikube installation - basic auth)

This example shows how to connect to a local minikube Kubernetes cluster which utilizes basic auth authentication.

When using basic auth, you need to start the minikube as shown below.

$ cat users.csv
pass123,user1,developers

# Mount a share with a local users file
minikube mount /home/libcloud/users.csv:/var/lib/docker/users.csv

# Start miniube
minikube --extra-config=apiserver.basic-auth-file=/var/lib/docker/users.csv start

import libcloud.security
from libcloud.container.types import Provider
from libcloud.container.providers import get_driver

# Disable cert vertification when running minikube locally using self signed
# cert
libcloud.security.VERIFY_SSL_CERT = False

cls = get_driver(Provider.KUBERNETES)

# You can retrieve cluster ip by running "minikube ip" command
conn = cls(key="user1", secret="pass123", host="192.168.99.100", port=8443, secure=True)

for container in conn.list_containers():
    print(container.name)

for cluster in conn.list_clusters():
    print(cluster.name)

Docker Hub Client HubClient is a shared utility class for interfacing to the public Docker Hub Service.

You can use this class for fetching images to deploy to services like ECS

from libcloud.container.types import Provider
from libcloud.container.providers import get_driver
from libcloud.container.utils.docker import HubClient

cls = get_driver(Provider.KUBERNETES)

conn = cls(key="my_username", secret="THIS_IS)+_MY_SECRET_KEY+I6TVkv68o4H", host="126.32.21.4")
hub = HubClient()

image = hub.get_image("ubuntu", "latest")

for cluster in conn.list_clusters():
    print(cluster.name)
    if cluster.name == "default":
        container = conn.deploy_container(cluster=cluster, name="my-simple-app", image=image)

API Docs

class libcloud.container.drivers.kubernetes.KubernetesContainerDriver(key=None, secret=None, secure=False, host='localhost', port=4243, key_file=None, cert_file=None, ca_cert=None, ex_token_bearer_auth=False)[source]
Parameters:
  • key (str) – API key or username to be used (required)
  • secret (str) – Secret password to be used (required)
  • secure (bool) – Whether to use HTTPS or HTTP. Note: Some providers only support HTTPS, and it is on by default.
  • host (str) – Override hostname used for connections.
  • port (int) – Override port used for connections.
  • key_file (str) – Path to the key file used to authenticate (when using key file auth).
  • cert_file (str) – Path to the cert file used to authenticate (when using key file auth).
  • ex_token_bearer_auth (bool) – True to use token bearer auth.
Returns:

None
connectionCls

alias of libcloud.common.kubernetes.KubernetesBasicAuthConnection

create_cluster(name, location=None)

Create a container cluster

Parameters:
  • name (str) – The name of the cluster
  • location (ClusterLocation) – The location to create the cluster in
Return type:

ContainerCluster
create_namespace(name: str) → libcloud.container.drivers.kubernetes.KubernetesNamespace[source]

Create a namespace

Parameters:name (str) – The name of the namespace
Return type:KubernetesNamespace
delete_namespace(namespace: libcloud.container.drivers.kubernetes.KubernetesNamespace) → bool[source]

Delete a namespace

Returns:True if the destroy was successful, otherwise False.
Return type:bool
deploy_container(name: str, image: libcloud.container.base.ContainerImage, namespace: libcloud.container.drivers.kubernetes.KubernetesNamespace = None, parameters: Optional[str] = None, start: Optional[bool] = True)[source]

Deploy an installed container image. In kubernetes this deploys a single container Pod. https://cloud.google.com/container-engine/docs/pods/single-container

Parameters:
  • name (str) – The name of the new container
  • image (ContainerImage) – The container image to deploy
  • namespace (KubernetesNamespace) – The namespace to deploy to, None is default
  • parameters (str) – Container Image parameters(unused)
  • start (bool) – Start the container on deployment(unused)
Return type:

Container
destroy_cluster(cluster)

Delete a cluster

Returns:True if the destroy was successful, otherwise False.
Return type:bool
destroy_container(container: libcloud.container.base.Container) → bool[source]

Destroy a deployed container. Because the containers are single container pods, this will delete the pod.

Parameters:container (Container) – The container to destroy
Return type:bool
ex_destroy_node(node_name: str) → bool[source]

Destroy a node.

Parameters:node_name (str) – Name of the node to destroy
Return type:bool
ex_destroy_pod(namespace: str, pod_name: str) → bool[source]

Delete a pod and the containers within it.

Parameters:
  • namespace (str) – The pod’s namespace
  • pod_name (str) – Name of the pod to destroy
Return type:

bool
ex_get_version() → str[source]

Get Kubernetes version

Return type:str
ex_list_deployments() → List[libcloud.container.drivers.kubernetes.KubernetesDeployment][source]

Get cluster deployments

Return type:list of KubernetesDeployment
ex_list_nodes() → List[libcloud.compute.base.Node][source]

List available Nodes

Return type:list of Node
ex_list_nodes_metrics() → List[Dict[str, Any]][source]

Get nodes metrics from Kubernetes Metrics Server

Return type:list of dict
ex_list_pods(fetch_metrics: bool = False) → List[libcloud.container.drivers.kubernetes.KubernetesPod][source]

List available Pods

Parameters:fetch_metrics (bool) – Fetch metrics for pods
Return type:list of KubernetesPod
ex_list_pods_metrics() → List[Dict[str, Any]][source]

Get pods metrics from Kubernetes Metrics Server

Return type:list of dict
ex_list_services() → List[Dict[str, Any]][source]

Get cluster services

Return type:list of dict
get_cluster(id)

Get a cluster by ID

Parameters:id (str) – The ID of the cluster to get
Return type:ContainerCluster
get_container(id: str) → libcloud.container.base.Container[source]

Get a container by ID

Parameters:id (str) – The ID of the container to get
Return type:libcloud.container.base.Container
get_namespace(id: str) → libcloud.container.drivers.kubernetes.KubernetesNamespace[source]

Get a namespace by ID

Parameters:id (str) – The ID of the namespace to get
Return type:KubernetesNamespace
install_image(path)

Install a container image from a remote path.

Parameters:path (str) – Path to the container image
Return type:ContainerImage
list_clusters(location=None)

Get a list of potential locations to deploy clusters into

Parameters:location (ClusterLocation) – The location to search in
Return type:list of ContainerCluster
list_containers(image=None, all=True) → List[libcloud.container.base.Container][source]

List the deployed container images

Parameters:
Return type:

list of libcloud.container.base.Container
list_images()

List the installed container images

Return type:list of ContainerImage
list_locations()

Get a list of potential locations to deploy clusters into

Return type:list of ClusterLocation
list_namespaces() → List[libcloud.container.drivers.kubernetes.KubernetesNamespace][source]

Get a list of namespaces that pods can be deployed into

Return type:list of KubernetesNamespace
restart_container(container)

Restart a deployed container

Parameters:container (Container) – The container to restart
Return type:Container
start_container(container)

Start a deployed container

Parameters:container (Container) – The container to start
Return type:Container
stop_container(container)

Stop a deployed container

Parameters:container (Container) – The container to stop
Return type:Container