Question 103848 in Azure Question posted in
The official documentation to get you started can be found here.

Populate pods CPU limits using Kubernetes Python Client for Azure AKS cluster

I need to use Azure Python SDK and Kubernetes Python Client to list the Pods CPU limits for a cluster running in AKS.

Although its straight forward using CLI/PowerShell but I need to use Python exclusively.
Must not use subprocess calls.

Here is snippet that gets KubeConfig object after authentication with Azure:

from azure.identity import DefaultAzureCredential
from azure.mgmt.containerservice import ContainerServiceClient

credential = DefaultAzureCredential(exclude_cli_credential=True)
subscription_id = "XXX"
resource_group_name= 'MY-SUB'
cluster_name = "my-aks-clustername" 
container_service_client = ContainerServiceClient(credential, subscription_id)

kubeconfig = container_service_client.managed_clusters. 
list_cluster_user_credentials(resource_group_name, cluster_name). 
kubeconfigs[0]

But I am unsure how to put this to be used by K8s Python client:

from kubernetes import client, config
config.load_kube_config() ## How to pass? 

v1 = client.CoreV1Api()
print("Listing pods with their IPs:")
ret = v1.list_pod_for_all_namespaces(watch=False)
for i in ret.items:
    print("%st%st%s" % (i.status.pod_ip, i.metadata.namespace, i.metadata.name))
Tags:

3

Answers


  1. 0 
    from kubernetes import client, config
config.load_kube_config() ## How to pass? 

v1 = client.CoreV1Api()
print("Listing pods with their CPU limits:")
ret = v1.list_pod_for_all_namespaces(watch=False)
for i in ret.items:
    for container in i.spec.containers:
        if container.resources.limits:
            if 'cpu' in container.resources.limits:
                print(container.resources.limits['cpu'])
    Login or Signup to reply.
  2. 0

    You can use the config.load_kube_config method and pass in the kubeconfig object you obtained earlier as a parameter. The method accepts a config_file parameter, which can be a file object, a file-like object, or a string file path.

    Since kubeconfig is a string, you can pass it as a string file path, like so:

    from kubernetes import client, config

# Pass the kubeconfig string as a file path
config.load_kube_config(config_file=kubeconfig)

v1 = client.CoreV1Api()
print("Listing pods with their IPs:")
ret = v1.list_pod_for_all_namespaces(watch=False)
for i in ret.items:
    print("%st%st%s" % (i.status.pod_ip, i.metadata.namespace, i.metadata.name))
    Login or Signup to reply.
  3. 0

    I tried a repro and was able to achieve as below –

    1. Getting kubeconfig in correct decoded format from kubeconfigs[0] which is CredentialResults.

    2. Writing the generated kubeconfig in a file.

    3. Loading the file in config module.

      from azure.identity import DefaultAzureCredential
from azure.mgmt.containerservice import ContainerServiceClient
import os
from kubernetes import client, config

credential = DefaultAzureCredential(exclude_cli_credential=True)
subscription_id = "XXX"
resource_group_name= 'MY-SUB'
cluster_name = "my-aks-clustername" 
container_service_client = ContainerServiceClient(credential, subscription_id)

# getting kubeconfig in a decoded format from CredentialResult

kubeconfig = container_service_client.managed_clusters. 
list_cluster_user_credentials(resource_group_name, cluster_name). 
kubeconfigs[0].value.decode(encoding='UTF-8')

# writing generated kubeconfig in a file
f=open("kubeconfig","w")
f.write(kubeconfig)
f.close()

# loading the config file
config.load_kube_config('kubeconfig')

# deleting the kubeconfig file
os.remove('kubeconfig')

v1 = client.CoreV1Api()
print("Listing containers with their CPU limits:")
ret = v1.list_pod_for_all_namespaces(watch=False)
for i in ret.items:
  for container in i.spec.containers:
    if container.resources.limits:
      if 'cpu' in container.resources.limits:
         print( container.name, container.resources.limits['cpu'])
    Login or Signup to reply.
Please signup or login to give your own answer.

Back To Top
Search