new_samples

Cadence Samples

These are some samples to demonstrate various capabilities of Cadence client and server. You can learn more about cadence at:

• The website: https://cadenceworkflow.io
• Cadence's server: https://github.com/uber/cadence
• Cadence's Go client: https://github.com/uber-go/cadence-client

Prerequisite

1. Install Cadence CLI. See instruction here.
2. Run the Cadence server: https://github.com/uber/cadence/blob/master/README.md
3. Register the cadence-samples domain: https://cadenceworkflow.io/docs/cli/#quick-start or https://github.com/uber/cadence/blob/master/tools/cli/README.md

Detailed workflow histories can be viewed on Cadence UI. Go to localhost:8088 on your browser for more details.

Steps to run samples

Simply start the go binary by running

go run main.go

Then you may interact with the samples using Cadence CLI. You may start multiple workers by running this binary. Here are a list of example CLIs you may use to play with the samples.

HelloWorld workflow

This workflow takes an input message and greet you as response. Try the following CLI

cadence --env development \
  --domain cadence-samples \
  workflow start \
  --workflow_type cadence_samples.HelloWorldWorkflow \
  --tl cadence-samples-worker \
  --et 60 \
  --input '{"message":"Uber"}'

Try to inspect the log message for the output.

Signal Workflow

This workflow demonstrate a basic signal workflow. It takes your name as input parameter and greet you based on languages you pick. To start the workflow, try the following CLI.

cadence --env development \
  --domain cadence-samples \
  workflow start \
  --workflow_type cadence_samples.SignalGreeterMultiLanguageWorkflow \
  --tl cadence-samples-worker \
  --et 6000 \
  --input '{"name":"Uber"}'

A workflow ID and run ID will be return in your terminal. Copy the workflow ID and replace to the CLI below to trigger the signal. Try to change the input language value and inspect what happens in the log. Also, try to inspect what happened after you interact with the signal multiple times.

cadence --env development \
  --domain cadence-samples \
  workflow signal \
  --workflow_id <Your workflow ID> \
  --name "language" \
  --input '"english"'

To cancel the workflow, try the following CLI.

cadence --env development \
  --domain cadence-samples \
  workflow signal \
  --workflow_id <Your workflow ID> \
  --name "cancel" \
  --input 'true'

The workflow should have a status of fail.

Dynamic workflow

This dynamic workflow is very similar to the Hello World workflow above, but instead of passing the function definition to the activity execution step, we can directly pass an activity name registered to the worker to run the workflow. Here is a sample CLI.

cadence --env development \
  --domain cadence-samples \
  workflow start \
  --workflow_type cadence_samples.DynamicWorkflow \
  --tl cadence-samples-worker \
  --et 60 \
  --input '{"message":"Uber"}'

Try to inspect the log message for the output.

Parallel pick first workflow

This example runs two activities in parallel and output the first successful response. Change the configuration for WaitForCancellation in the activity options will change the behavior of canceling other unfinished activities. Here is a sample CLI

cadence --env development \
  --domain cadence-samples \
  workflow start \
  --workflow_type cadence_samples.ParallelBranchPickFirstWorkflow \
  --tl cadence-samples-worker \
  --et 60 \
  --input '{}'

Cancel workflow

This workflow includes multiple activities and each will respond to the workflow cancel action differently. To start the workflow first, simply run

cadence --env development \
  --domain cadence-samples \
  workflow start \
  --workflow_type cadence_samples.CancelWorkflow \
  --tl cadence-samples-worker \
  --et 60 \
  --input '{}'

You should have a workflow ID output in the terminal. Then to cancel the workflow, run

cadence --env development \
  --domain cadence-samples \
  workflow cancel \
  --workflow_id <YOUR_WORKFLOW_ID>

Inspect the behavior of the activities in the console.