Skip to content
This repository has been archived by the owner on Sep 18, 2023. It is now read-only.


Repository files navigation

Intro: Fault-Oblivious Stateful Python Code

cadence-python allows you to create Python functions that have their state (local variables etc..) implicitly saved such that if the process/machine fails the state of the function is not lost and can resume from where it left off.

This programming model is useful whenever you need to ensure that a function runs to completion. For example:

  • Business logic involving multiple micro services
  • CI/CD pipelines
  • Data pipelines
  • RPA
  • ETL
  • Marketing automation / Customer journeys / Customer engagement
  • Zapier/IFTTT like end user automation.
  • Chat bots
  • Multi-step forms
  • Scheduler/Cron jobs

Behind the scenes, cadence-python uses Cadence as its backend.

For more information about the fault-oblivious programming model refer to the Cadence documentation here

Install Cadencce

docker-compose up

Register sample domain

docker run --network=host --rm ubercadence/cli:master --do sample domain register -rd 1

Installation cadence-python

pip install cadence-client==1.0.1

Hello World Sample

import sys
import logging
from cadence.activity_method import activity_method
from cadence.workerfactory import WorkerFactory
from cadence.workflow import workflow_method, Workflow, WorkflowClient


TASK_LIST = "HelloActivity-python-tasklist"
DOMAIN = "sample"

# Activities Interface
class GreetingActivities:
    @activity_method(task_list=TASK_LIST, schedule_to_close_timeout_seconds=2)
    def compose_greeting(self, greeting: str, name: str) -> str:
        raise NotImplementedError

# Activities Implementation
class GreetingActivitiesImpl:
    def compose_greeting(self, greeting: str, name: str):
        return f"{greeting} {name}!"

# Workflow Interface
class GreetingWorkflow:
    @workflow_method(execution_start_to_close_timeout_seconds=10, task_list=TASK_LIST)
    async def get_greeting(self, name: str) -> str:
        raise NotImplementedError

# Workflow Implementation
class GreetingWorkflowImpl(GreetingWorkflow):

    def __init__(self):
        self.greeting_activities: GreetingActivities = Workflow.new_activity_stub(GreetingActivities)

    async def get_greeting(self, name):
        # Place any Python code here that you want to ensure is executed to completion.
        # Note: code in workflow functions must be deterministic so that the same code paths
        # are ran during replay.
        return await self.greeting_activities.compose_greeting("Hello", name)

if __name__ == '__main__':
    factory = WorkerFactory("localhost", 7933, DOMAIN)
    worker = factory.new_worker(TASK_LIST)
    worker.register_activities_implementation(GreetingActivitiesImpl(), "GreetingActivities")

    client = WorkflowClient.new_client(domain=DOMAIN)
    greeting_workflow: GreetingWorkflow = client.new_workflow_stub(GreetingWorkflow)
    result = greeting_workflow.get_greeting("Python")

    print("Stopping workers....")
    print("Workers stopped...")

Status / TODO

cadence-python is still under going heavy development. It should be considered EXPERIMENTAL at the moment. A production version is targeted to be released in September of 2019 January 2020 March 2020 April 2020.


  • Tchannel implementation
  • Python-friendly wrapper around Cadence's Thrift API
  • Author activities in Python
  • Start workflows (synchronously)
  • Create workflows
  • Workflow execution in coroutines
  • Invoke activities from workflows
  • ActivityCompletionClient heartbeat, complete, complete_exceptionally
  • Activity heartbeat, getHeartbeatDetails and doNotCompleteOnReturn
  • Activity retry
  • Activity getDomain(), getTaskToken(), getWorkflowExecution()
  • Signals
  • Queries
  • Async workflow execution
  • await
  • now (currentTimeMillis)
  • Sleep
  • Loggers
  • newRandom
  • UUID
  • Workflow Versioning
  • WorkflowClient.newWorkflowStub(Class workflowInterface, String workflowId);


  • ActivityStub and Workflow.newUntypedActivityStub
  • Classes as arguments and return values to/from activity and workflow methods
  • WorkflowStub and WorkflowClient.newUntypedWorkflowStub
  • Custom workflow ids through start() and new_workflow_stub()
  • ContinueAsNew
  • Compatibility with Java client
  • Compatibility with Golang client


  • Sticky workflows

Post 2.0:

  • sideEffect/mutableSideEffect
  • Local activity
  • Parallel activity execution
  • Timers
  • Cancellation Scopes
  • Child Workflows
  • Explicit activity ids for activity invocations