Skip to main content

Flow Run Automation

Programmatically Running Flows​

Flows can be run programmatically using the GanymedeFlowRun object in ganymede_sdk.flow_run.

GanymedeFlowRun requires API configuration and a Flow ID to run a Flow. Reach out to Ganymede Support for API configuration assistance if needed.

In the example below, an API key called "ganymede_api_key_demo" is configured in the environment. A string with value 'EXP001' is attached to the Input_Param node, and a CSV file is attached to the CSV_Read node in a Flow with an ID of demo-flow.

from ganymede_sdk.flow_runtime import GanymedeFlowRun
from ganymede_sdk import get_secret

flow_run = GanymedeFlowRun(api_key=get_secret("ganymede_api_key_demo"))

flow_run.attach_param(node='Input_Param', value='EXP001')
flow_run.attach_file(node="CSV_Read", param="csv", files={"input_file.txt": b"Hello World"})

# Trigger the Flow Run using the parameters attached to the GanymedeFlowRun object
flow_run.trigger_flow_run("demo-flow")

To monitor the status of the Flow Run, use the get_flow_run_status method.

# retrieve Flow run status for the latest execution of trigger_flow_run
flow_run.get_flow_run_status()

# retrieve Flow run status for a specific Flow run
flow_run.get_flow_run_status(flow_name="demo-flow", run_id="1234567890")

If you would like to poll for Flow run completion, use the wait_for_flow_run_completion method.

flow_run.wait_for_flow_run_completion()

method attach_file​

Upload a file to Ganymede and associate with Node for Flow run. The uploaded file will be stored in the Ganymede input bucket as _api/{file_name}.

  • param node : str - Name of the node
  • param param : str - Name of the parameter (typically, file extension associated with the node)
  • param files : dict[str, bytes | BytesIO] - Dictionary keyed by the name of the file(s) to upload, with values as the file contents
  • param is_multi : bool, optional - Whether input node references single or multiple files. Defaults to False.

method attach_param​

Attach an input string parameter to the Flow run

  • param node : str - Name of the node
  • param value : str - Value of the parameter

method reference_file​

Reference a file in the Ganymede input bucket to a node for Flow run.

  • param node : str - Name of the node
  • param param : str - Name of the parameter (typically, file extension associated with the node)
  • param file_name : str | list[str] - Name of the file in the Ganymede input bucket
  • param is_multi: bool, optional - Whether input node references single or multiple files. Defaults to False.

method trigger_flow_run​

Trigger a Flow run with the attached parameters. Returns a RunConf object containing the Flow run configuration.

  • param flow : str - ID of the Flow to run
  • param run_conf: RunConf | None - Run configuration for the Flow run if specified. If not specified, a new run configuration will be created based on calls to attach_param, attach_file, and reference_file.

method get_flow_run_status​

Obtain Flow run status for specified flow and run ID. If not specified, retrieve status for latest execution of trigger_flow_run.

  • param flow_name : str, optional - Name of the Flow to retrieve status for. If not specified, retrieve status for latest execution of trigger_flow_run.
  • param run_id : str, optional - Run ID of the Flow run to retrieve status for. If not specified, retrieve run ID for latest execution of trigger_flow_run.

method wait_for_flow_run_completion_async​

Asynchronously polls the get_flow_run_status endpoint every check_interval seconds until the Flow is in a terminal state (Success/Failed). Returns the final status dictionary.

  • param flow_name : str, optional - Name of the Flow to retrieve status for. If not specified, retrieve status for latest execution of trigger_flow_run.
  • param run_id : str, optional - Run ID of the Flow run to retrieve status for. If not specified, retrieve run ID for latest execution of trigger_flow_run.
  • param check_interval : int, optional - Interval in seconds to poll for Flow status. Defaults to 5 seconds.
  • param timeout_seconds : int, optional - Timeout in seconds to wait for Flow completion. Defaults to 3600 seconds (1 hour).
  • param verbose : bool, optional - Whether to print status updates to console at each check interval. Defaults to False.

method wait_for_flow_run_completion​

Synchronously polls for flow run completion by running the async method in an event loop. This method can be run in either a Jupyter notebook or a Python script.

  • param flow_name : str, optional - Name of the Flow to retrieve status for. If not specified, retrieve status for latest execution of trigger_flow_run.
  • param run_id : str, optional - Run ID of the Flow run to retrieve status for. If not specified, retrieve run ID for latest execution of trigger_flow_run.
  • param check_interval : int, optional - Interval in seconds to poll for Flow status. Defaults to 5 seconds.
  • param timeout_seconds : int, optional - Timeout in seconds to wait for Flow completion. Defaults to 3600 seconds (1 hour).
  • param verbose : bool, optional - Whether to print status updates to console at each check interval. Defaults to False.