All the samples are self contained unless they are placed inside their own folders. The samples use Application Default Credentails (ADC) to authenticate with GCP. So make sure ADC is setup correctly (i.e. GOOGLE_APPLICATION_CREDENTIALS
environment variable is set) before running the samples. Some sample might require additional python modules to be installed.
You can run samples as follows:
python <sample_name.py> <arg1> <arg2> ...
You can run the following command to find the usage and arguments for the samples:
python <sample_name.py> -h
# example
python quickstart.py -h
usage: quickstart.py [-h] project_id zone
positional arguments:
project_id Google Cloud project ID
zone GKE Cluster zone
optional arguments:
-h, --help show this help message and exit
- quickstart.py: A simple example to list the GKE clusters in a given GCP project and zone. The sample uses the
list_clusters()
API to fetch the list of cluster.
The following samples are examples of operations that take a while to complete.
For example creating a cluster in GKE can take a while to set up the cluster
nodes, networking and configuring Kubernetes. Thus, calls to such long running
APIs return an object of type Operation
. We can
then use the id of the returned operation to poll the get_operation()
API to check for it's status. You can see the
different statuses it can be in, in this proto definition.
-
create_cluster.py: An example of creating a GKE cluster (with mostly the defaults). This example shows how to handle responses of type
Operation
that reperesents a long running operation. The example uses the python modulebackoff
to handle a graceful exponential backoff retry mechanism to check if theOperation
has completed. -
delete_cluster.py: An example of deleting a GKE cluster. This example shows how to handle responses of type
Operation
that reperesents a long running operation.