End to End Testing in Kubernetes – Part II

In this blog post, I will show you how to write End to End Tests in Kubernetes to verify the various objects in a Kubernetes cluster like Pods, Services, Nodes, Namespaces etc.

I would highly recommend you to read my previous blog about End to End Testing in Kubernetes for context setting prior to looking at the below set of tests – 

End to End Testing in Kubernetes

Health Check
If you want to interact with the Kubernetes API, there is no need to write your own library. You can leverage a number of available client libraries for this purpose. I am using the Python API Client for Kubernetes to interact with the cluster in Cloud.

Official Python client library for Kubernetes –
https://github.com/kubernetes-client/python

Documentation for Kubernetes API Endpoints –
https://github.com/kubernetes-client/python/tree/master/kubernetes#documentation-for-api-endpoints

The below tests will verify the pod count and service count in the ‘default’ namespace —

def test_to_verify_pod_count_in_default_namespace(k8s_client):
podcount = len(k8s_client.list_namespaced_pod(namespace='default').items)
print(podcount)
assert(podcount > 0)
def test_to_verify_service_count_in_default_namespace(k8s_client):
svccount = len(k8s_client.list_namespaced_service(namespace='default').items)
print(svccount)
assert(svccount > 0)

The below test will iterate through all the namespaces in the K8s cluster and list it down –

def test_to_list_all_namespaces(k8s_client):
ret = k8s_client.list_namespace()
nscount = len(ret.items)
print(nscount)
assert(nscount > 0)
# Iterate through all the namespaces in the K8s cluster and list it down
for item in ret.items:
print(item.metadata.name)

The below set of tests will iterate through all the endpoints, services, nodes and pods in the K8s cluster and list them down —

def test_to_list_endpoints_for_all_namespaces(k8s_client):
ret = k8s_client.list_endpoints_for_all_namespaces()
epcount = len(ret.items)
print(epcount)
assert(epcount > 0)
# Iterate through all the endpoints in the K8s cluster and list it down
for item in ret.items:
print("%s\t%s\t" % (item.metadata.name, item.metadata.namespace))
def test_to_list_services_for_all_namespaces(k8s_client):
ret = k8s_client.list_service_for_all_namespaces()
print("\n")
svccount = len(ret.items)
print(svccount)
assert(svccount > 0)
# Iterate through all the services in the K8s cluster and list down all the services and respective namespace
for item in ret.items:
print("%s\t%s\t" % (item.metadata.name, item.metadata.namespace))
def test_to_list_nodes_for_all_namespaces(k8s_client):
ret = k8s_client.list_node()
nodecount = len(ret.items)
print(nodecount)
assert(nodecount > 0)
# Iterate through all the nodes in the K8s cluster and list it down
for item in ret.items:
print("%s\t" % item.metadata.name)
def test_to_list_pods_for_all_namespaces(k8s_client):
ret = k8s_client.list_pod_for_all_namespaces()
podcount = len(ret.items)
print(podcount)
assert(podcount > 0)
# Iterate through all the pods in the K8s cluster and list down the pod name and respective namespace
for item in ret.items:
print("%s\t%s\t" % (item.metadata.name, item.metadata.namespace))

The below tests lists pods and services for a specific namespace —

def test_to_list_pods_for_specific_namespace(k8s_client):
ret = k8s_client.list_namespaced_pod("default")
podcount = len(ret.items)
print(podcount)
assert(podcount > 0)
# Iterate through all the pods in the default namespace and list down the pod name and respective namespace
for item in ret.items:
print("%s\t%s\t" % (item.metadata.name, item.metadata.namespace))
def test_to_list_services_for_specific_namespace(k8s_client):
ret = k8s_client.list_namespaced_service("default")
svccount = len(ret.items)
print(svccount)
assert(svccount > 0)
# Iterate through all the services in the default namespace and list down the service name and respective namespace
for item in ret.items:
print("%s\t%s\t" % (item.metadata.name, item.metadata.namespace))

Using the Python API Client we saw how to interact with the K8s components in the cluster hosted in AWS or Azure. By running a set of End to End Tests we can easily determine the health of the cluster. We have just scratched the surface with these E2E Tests — you can improvise on top of these tests to interact with any K8s object in the cluster and verify functionalities.



Categories: Kubernetes

4 replies

  1. is there any python e2e test framework like e2e(ginkgo) in golang?

Trackbacks

  1. End to End Testing in Kubernetes – dotnetvibes
  2. Crash Loop Detection in Kubernetes – dotnetvibes
  3. Kubernetes Architecture – dotnetvibes

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: