def find_experiment_id(experiment_name: str, client: kfp.Client, page_size: int = 100, page_token: str = "") -> str: """Function to return the experiment id Arguments: experiment_name {str} -- The experiment name client {kfp.Client} -- The kfp client Returns: str -- The experiment id """ while True: experiments = client.list_experiments(page_size=page_size, page_token=page_token) for experiments in experiments.experiments: if experiments.name == experiment_name: logging.info("Succesfully collected the experiment id") return experiments.id # Start need to know where to do next itteration from page_token = experiments.next_page_token # If no next tooken break if not page_token: logging.info( f"Could not find the pipeline id, is the experiment name: {experiments_name} correct? " ) break
def get_or_create_experiment(experiment_name: str, client: kfp.Client) -> ApiExperiment: existing_experiments = client.list_experiments().experiments if existing_experiments is not None: exp = next(iter([exp for exp in existing_experiments if exp.name == experiment_name]), None) else: exp = None if exp is None: exp = client.create_experiment(experiment_name) print('Experiment %s created with ID %s' % (exp.name, exp.id)) else: print('Experiment already exists with id %s' % exp.id) return exp
def experiment_id(client: kfp.Client, name: str): """Gets the ID of the kubeflow experiment with the name 'name' Args: name of the experiment Returns: id of the experiment """ page_token = "" while page_token is not None: p = client.list_experiments(page_token=page_token, page_size=100) if p.experiments is None: return "" for p in p.experiments: if p.name == name: return p.id page_token = p.next_page_token return ""
def main(): args = parse_arguments() test_cases = [] test_name = args.testname + ' Sample Test' ###### Initialization ###### client = Client(namespace=args.namespace) ###### Get experiments ###### list_experiments_response = client.list_experiments(page_size=100) for experiment in list_experiments_response.experiments: if experiment.name == args.experiment: experiment_id = experiment.id ###### Get runs ###### import kfp_run resource_reference_key_type = kfp_run.models.api_resource_type.ApiResourceType.EXPERIMENT resource_reference_key_id = experiment_id list_runs_response = client.list_runs( page_size=1000, resource_reference_key_type=resource_reference_key_type, resource_reference_key_id=resource_reference_key_id) ###### Check all runs ###### for run in list_runs_response.runs: run_id = run.id response = client.wait_for_run_completion(run_id, 1200) succ = (response.run.status.lower() == 'succeeded') utils.add_junit_test(test_cases, 'job completion', succ, 'waiting for job completion failure') ###### Output Argo Log for Debugging ###### workflow_json = client._get_workflow_json(run_id) workflow_id = workflow_json['metadata']['name'] argo_log, _ = utils.run_bash_command('argo logs -n {} -w {}'.format( args.namespace, workflow_id)) print("=========Argo Workflow Log=========") print(argo_log) if not succ: utils.write_junit_xml(test_name, args.result, test_cases) exit(1) ###### Write out the test result in junit xml ###### utils.write_junit_xml(test_name, args.result, test_cases)
class KubeflowClient: """ A wrapper of the existing Kubeflow Pipelines Client which enriches it to be able to access more of the Kubeflow Pipelines API. """ def __init__(self, host: Optional[str] = None, client_id: Optional[str] = None, namespace: Optional[str] = "kubeflow"): """ Instandiate a new KubeflowClient Args: host (str): The host we can find the Kubeflow API at (e.g. https://{APP_NAME}.endpoints.{PROJECT_ID}.cloud.goog/pipeline) client_id (str): The IAP client id we can use for authorisate (e.g. "XXXXXX-XXXXXXXXX.apps.googleusercontent.com") namespace (str): The Kuberenetes / Kubeflow namespace to deploy to (e.g. kubeflow) """ self.host = host self.client_id = client_id self.namespace = namespace logging.info(f"KubeflowClient: host: {host}, client_id: {client_id}") self.kfp_client = Client(host, client_id, namespace) self.config = self.kfp_client._load_config(self.host, self.client_id, self.namespace, None, None) # print(f"kfp auth:") # print(f"\thost: {self.host}") # print(f"\tclient_id: {self.client_id}") # print(f"\tnamespace: {self.namespace}") # print(f"\tapi_key: {self.config.api_key}") self.kfp_pipelines = self._connect_pipelines_api() self.kfp_runs = self._connect_runs_api() self.kfp_jobs = self._connect_jobs_api() def create_pipeline(self, pipeline_func, pipeline_name): """ Create a new Kubeflow Pipeline using the provided pipeline function Args: pipeline_func: The method decorated by @dsl.pipeline which defines the pipeline Returns: The Kubeflow Pipeline object created """ try: (_, pipeline_package_path) = tempfile.mkstemp(suffix=".zip") compiler.Compiler().compile(pipeline_func, pipeline_package_path) logging.info(f"Compiled piopeline to: {pipeline_package_path}") return self.kfp_client.upload_pipeline(pipeline_package_path, pipeline_name) finally: pass # os.remove(pipeline_package_path) def create_experiment(self, experiment_name): """ Create a new Kubeflow Pipelines Experiment (grouping of pipeliens / runs) Args: experiment_name (str): The name of the experiment Returns: The Kubeflow experiement object created """ return self.kfp_client.create_experiment(name=experiment_name) def list_experiments(self): """ List the Experiments in the current namespace Returns: A list of all the Experiments """ all_experiments = list() next_page_token = "" while next_page_token is not None: response = self.kfp_client.list_experiments( page_size=100, page_token=next_page_token) if response.experiments is None: break all_experiments.extend(response.experiments) next_page_token = response.next_page_token count = len(all_experiments) # print(f"list_experiments: found {count}") return all_experiments def find_job(self, job_name): """ Look up a job by its name (in the current namespace). Returns None if the job cannot be found Args: job_name (str): The name of the job to find Returns: A reference to the job if found, and None if not. """ jobs = self.list_jobs() if jobs is None: return None for j in jobs: if j.name == job_name: return j return None def list_jobs(self): """ List the Jobs in the current namespace Returns: A list of all the Jobs """ all_jobs = list() next_page_token = "" while next_page_token is not None: response = self.kfp_jobs.list_jobs(page_size=100, page_token=next_page_token) if response.jobs is None: break all_jobs.extend(response.jobs) next_page_token = response.next_page_token count = len(all_jobs) # print(f"all_jobs: found {count}") return all_jobs def delete_job(self, job): """ Delete a `Job` using its job.id Args: job (KubeflowJob): A `Job` object to delete Returns: True if the `Job` was deleted succesfully """ self.kfp_jobs.delete_job(id=job.id) return True def create_job(self, name: str, pipeline, experiment, description=None, enabled=True, max_concurrency=1, cron=None): """ Create a new Kubeflow Pipelines Job Args: name (str): The name of the `Job` pipeline (Pipeline): The `Pipeline` object to execute when the `Job` is called experiment (Experiment): The `Experiment` object to create the `Job` in. description (str): A description of what the `Job` is all about enabled (bool): Should be `Job` be enabled? max_concurrency (int): How many concurrent executions of the `Job` are allowed? cron (str): The CRON expression to use to execute the job periodicalls Returns: The Kubeflow API response object. """ key = kfp_server_api.models.ApiResourceKey( id=experiment.id, type=kfp_server_api.models.ApiResourceType.EXPERIMENT) reference = kfp_server_api.models.ApiResourceReference( key, kfp_server_api.models.ApiRelationship.OWNER) spec = kfp_server_api.models.ApiPipelineSpec(pipeline_id=pipeline.id) trigger = None if cron is not None: cron_schedule = kfp_server_api.models.api_cron_schedule.ApiCronSchedule( cron=cron) trigger = kfp_server_api.models.api_trigger.ApiTrigger( cron_schedule=cron_schedule) run_body = kfp_server_api.models.ApiJob( name=name, description=description, pipeline_spec=spec, resource_references=[reference], enabled=True, trigger=trigger, max_concurrency=str(max_concurrency), ) response = self.kfp_jobs.create_job(body=run_body) return response def list_runs(self, experiment_name): """ List the `Runs` in the specified Exper`iment Args: experiment_name (str): The name of the `Experiment` Returns: A list of all the `Runs` in the current `Experiment` """ experiment = self.get_experiment(experiment_name=experiment_name) all_runs = list() next_page_token = "" while next_page_token is not None: response = self.kfp_client.list_runs(page_size=100, page_token=next_page_token) if response.runs is None: break all_runs.extend(response.runs) next_page_token = response.next_page_token run_count = len(all_runs) # print(f"list_runs: found {run_count}") return all_runs def list_pipelines(self): """ List the `Pipelines` in the current namespace Returns: A list of all the `Pipelines` in the current `Experiment` """ all_pipelines = list() response = self.kfp_client.list_pipelines(page_size=100) next_page_token = "" while next_page_token is not None: response = self.kfp_client.list_pipelines( page_size=100, page_token=next_page_token) if response.pipelines is None: break all_pipelines.extend(response.pipelines) next_page_token = response.next_page_token pipeline_count = len(all_pipelines) # print(f"list_pipelines: found {pipeline_count}") return all_pipelines def find_experiment(self, id=None, name=None): """ Look up an `Experiment` by its name or id. Returns None if the `Experiment` cannot be found. Both `id` and `name` are optional, but atleast one must be provided. Where both a provided, the function will return with the first `Experiment` matching either id or name. Args: id (str): The `id` of the `Experiment` to find name (string): The `name` of the `Experiment` to find Returns: A reference to the `Experiment` if found, and None if not. """ experiments = self.list_experiments() if experiments is None: return None for e in experiments: if e.name == name: return e if e.id == id: return e return None def find_pipeline(self, name): """ Look up a `Pipeline` by its name (in the current namespace). Returns None if the `Pipeline` cannot be found Args: name (str): The name of the `Pipeline` to find Returns: A reference to the `Pipeline` if found, and `None` if not. """ pipelines = self.list_pipelines() if pipelines is None: return None for p in pipelines: if p.name == name: return p return None def delete_pipeline(self, pipeline): """ Delete the specified `Pipeline` Args: pipeline: The pipeline object to delete Returns: True if successfull """ # Go through all my pipelines to find the one to delete self.kfp_pipelines.delete_pipeline(pipeline.id) return True def _connect_pipelines_api(self): """ Create a new PipelineServiceApi client """ api_client = kfp_server_api.api_client.ApiClient(self.config) pipelines_api = kfp_server_api.api.pipeline_service_api.PipelineServiceApi( api_client) return pipelines_api def _connect_runs_api(self): """ Create a new PipelineServiceApi client """ api_client = kfp_server_api.api_client.ApiClient(self.config) runs_api = kfp_server_api.api.run_service_api.RunServiceApi(api_client) return runs_api def _connect_jobs_api(self): """ Create a new PipelineServiceApi client """ api_client = kfp_server_api.api_client.ApiClient(self.config) runs_api = kfp_server_api.api.job_service_api.JobServiceApi(api_client) return runs_api