I am using google-cloud-tasks library to create tasks via a small python program. That program will be running inside a dataflow worker node. The default behaviour of the library is that it searches for credentials locally. But, if not found, it will call gcp metadata server to obtain the credentials ephimerally. Now, I have written a code where I am extracting the credential file from a secret manager and storing it inside the worker machine. This was done to ensure that same functionality is implemented but w/o the dependency on the metadata server. The function looks like this:
if not os.environ.get("service_file", default=None):
print("service file not found.")
print("Creating one.")
with open("/tmp/service_file.json", "w") as fp:
json.dump(
json.loads(
access_secret_manager_data(
service_cred_url
)
),
fp
)
os.environ["service_file"] = "/tmp/service_file.json"
# The code below sets the actual env variable in use.
os.environ["GOOGLE_APPLICATION_CREDENTIALS"] = "/tmp/service_file.json"
print(f"service file set. The location is: {os.environ['GOOGLE_APPLICATION_CREDENTIALS']}")
else:
print(f"service file present. The location is: {os.environ['GOOGLE_APPLICATION_CREDENTIALS']}")
# Below is the code where the client searches for the env variable we set above.
task_client = tasks_v2.CloudTasksClient()
task_client.create_task(...)
My query: is there a way to check via test libraries whether the cloud task library is using the env variable I have set above? I am new to testing and hence need some guidance.
Any information will be helpful.
