Execute a workflow using the Cloud Client Libraries

Deploy a sample workflow

After defining a workflow, you deploy it to make it available for execution. The deploy step also validates that the source file can be executed.

The following workflow sends a request to a public API and then returns the API's response.

1. Create a text file with the filename myFirstWorkflow.yaml with the following content:

   # This workflow accepts an optional "searchTerm" argument for the Wikipedia API.
   # If no input arguments are provided or "searchTerm" is absent,
   # it will fetch the day of the week in Amsterdam and use it as the search term.
   
   main:
       params: [input]
       steps:
       - validateSearchTermAndRedirectToReadWikipedia:
           switch:
               - condition: '${map.get(input, "searchTerm") != null}'
                 assign:
                   - searchTerm: '${input.searchTerm}'
                 next: readWikipedia
       - getCurrentTime:
           call: http.get
           args:
               url: https://timeapi.io/api/Time/current/zone?timeZone=Europe/Amsterdam
           result: currentTime
       - setFromCallResult:
           assign:
               - searchTerm: '${currentTime.body.dayOfWeek}'
       - readWikipedia:
           call: http.get
           args:
               url: 'https://en.wikipedia.org/w/api.php'
               query:
                   action: opensearch
                   search: '${searchTerm}'
           result: wikiResult
       - returnOutput:
               return: '${wikiResult.body[1]}'

2. After creating the workflow, you can deploy it, but don't execute the workflow:

   gcloud workflows deploy myFirstWorkflow \
       --source=myFirstWorkflow.yaml \
       --service-account=SERVICE_ACCOUNT_NAME@PROJECT_ID.iam.gserviceaccount.com \
       --location=CLOUD_REGION

   Replace CLOUD_REGION with a supported location for the workflow. The default region used in the code samples is us-central1.

Get the sample code

You can clone the sample code from GitHub.

1. Clone the sample app repository to your local machine:

   C#

   git clone https://github.com/GoogleCloudPlatform/dotnet-docs-samples.git

   Alternatively, you can download the sample as a zip file and extract it.

   Go

   git clone https://github.com/GoogleCloudPlatform/golang-samples.git

   Alternatively, you can download the sample as a zip file and extract it.

   Java

   git clone https://github.com/GoogleCloudPlatform/java-docs-samples.git

   Alternatively, you can download the sample as a zip file and extract it.

   Node.js

   git clone https://github.com/GoogleCloudPlatform/nodejs-docs-samples.git

   Alternatively, you can download the sample as a zip file and extract it.

   Python

   git clone https://github.com/GoogleCloudPlatform/python-docs-samples.git

   Alternatively, you can download the sample as a zip file and extract it.

2. Change to the directory that contains the Workflows sample code:

   C#

   cd dotnet-docs-samples/workflows/api/Workflow.Samples/

   Go

   cd golang-samples/workflows/executions/

   Java

   cd java-docs-samples/workflows/cloud-client/

   Node.js

   cd nodejs-docs-samples/workflows/quickstart/

   Python

   cd python-docs-samples/workflows/cloud-client/

3. Take a look at the sample code. Each sample app does the following:

   1. Sets up the Cloud Client Libraries for Workflows.
   2. Executes a workflow.
   3. Polls the workflow's execution (using exponential backoff) until the execution terminates.
   4. Prints the execution results.

   C#

   
   using Google.Cloud.Workflows.Common.V1;
   using Google.Cloud.Workflows.Executions.V1;
   using System;
   using System.Threading;
   using System.Threading.Tasks;
   
   public class ExecuteWorkflowSample
   {
       /// <summary>
       /// Execute a workflow and return the execution operation.
       /// </summary>
       /// <param name="projectID">Your Google Cloud Project ID.</param>
       /// <param name="locationID">The region where your workflow is located.</param>
       /// <param name="workflowID">Your Workflow ID.</param>
       /// <returns>
       /// An Execute object representing the completed workflow execution.
       /// </returns>
       public async Task<Execution> ExecuteWorkflow(
           string projectId = "YOUR-PROJECT-ID",
           string locationID = "YOUR-LOCATION-ID",
           string workflowID = "YOUR-WORKFLOW-ID")
       {
           // Initialize the client.
           ExecutionsClient client = await ExecutionsClient.CreateAsync();
   
           // Build the parent location path.
           WorkflowName parent = new WorkflowName(projectId, locationID, workflowID);
   
           // Create an execution request.
           CreateExecutionRequest createExecutionRequest = new CreateExecutionRequest
           {
               ParentAsWorkflowName = parent,
           };
   
           // Execute the operation.
           Execution execution = await client.CreateExecutionAsync(createExecutionRequest);
           Console.WriteLine("- Execution started...");
   
           TimeSpan backoffDelay = TimeSpan.FromSeconds(1);
           TimeSpan maxBackoffDelay = TimeSpan.FromSeconds(16);
   
           // Keep polling the state until the execution finishes, using exponential backoff.
           while (execution.State == Execution.Types.State.Active)
           {
               await Task.Delay(backoffDelay);
   
               // Implement exponential backoff by doubling the delay, but limiting it to a practical duration.
               backoffDelay = (backoffDelay < maxBackoffDelay) ? backoffDelay * 2 : maxBackoffDelay;
   
               execution = await client.GetExecutionAsync(execution.Name);
           }
   
           // Print results.
           Console.WriteLine($"Execution finished with state: {execution.State}");
           Console.WriteLine($"Execution results: {execution.Result}");
   
           // Return the fetched execution.
           return execution;
       }
   }

   Go

   import (
   	"context"
   	"fmt"
   	"io"
   	"time"
   
   	workflowexecutions "google.golang.org/api/workflowexecutions/v1"
   )
   
   // Execute a workflow and print the execution results.
   //
   // For more information about Workflows see:
   // https://cloud.google.com/workflows/docs/overview
   func executeWorkflow(w io.Writer, projectID, workflowID, locationID string) error {
   	// TODO(developer): Uncomment and update the following lines:
   	// projectID := "YOUR_PROJECT_ID"
   	// workflowID := "YOUR_WORKFLOW_ID"
   	// locationID := "YOUR_LOCATION_ID"
   
   	ctx := context.Background()
   
   	// Construct the location path.
   	parent := fmt.Sprintf("projects/%s/locations/%s/workflows/%s", projectID, locationID, workflowID)
   
   	// Create execution client.
   	client, err := workflowexecutions.NewService(ctx)
   	if err != nil {
   		return fmt.Errorf("workflowexecutions.NewService error: %w", err)
   	}
   
   	// Get execution service.
   	service := client.Projects.Locations.Workflows.Executions
   
   	// Build and run the new workflow execution.
   	res, err := service.Create(parent, &workflowexecutions.Execution{}).Do()
   	if err != nil {
   		return fmt.Errorf("service.Create.Do error: %w", err)
   	}
   	fmt.Fprintln(w, "- Execution started...")
   
   	// Set initial value for backoff delay in one second.
   	backoffDelay := time.Second
   
   	for res.State == "ACTIVE" {
   		time.Sleep(backoffDelay)
   
   		// Request the updated state for the execution.
   		getReq := service.Get(res.Name)
   		res, err = getReq.Do()
   		if err != nil {
   			return fmt.Errorf("getReq error: %w", err)
   		}
   
   		// Double the delay to provide exponential backoff (capped at 16 seconds).
   		if backoffDelay < time.Second*16 {
   			backoffDelay *= 2
   		}
   	}
   
   	fmt.Fprintf(w, "Execution finished with state: %s\n", res.State)
   	fmt.Fprintf(w, "Execution results: %s\n", res.Result)
   
   	return nil
   }
   

   Java

   
   import com.google.cloud.workflows.executions.v1.CreateExecutionRequest;
   import com.google.cloud.workflows.executions.v1.Execution;
   import com.google.cloud.workflows.executions.v1.ExecutionsClient;
   import com.google.cloud.workflows.executions.v1.WorkflowName;
   import java.io.IOException;
   import java.util.concurrent.ExecutionException;
   
   public class ExecuteWithoutArguments {
     public static void main(String[] args)
         throws IOException, InterruptedException, ExecutionException {
       String projectId = "your-project-id";
       String location = "your-location"; // For example: us-central1
       String workflowId = "your-workflow-id";
   
       executeWorkflowWithoutArguments(projectId, location, workflowId);
     }
   
     public static void executeWorkflowWithoutArguments(
         String projectId, String location, String workflowId)
         throws IOException, InterruptedException, ExecutionException {
       try (ExecutionsClient executionsClient = ExecutionsClient.create()) {
         WorkflowName parent = WorkflowName.of(projectId, location, workflowId);
   
         CreateExecutionRequest executionRequest =
             CreateExecutionRequest.newBuilder()
                 .setParent(parent.toString())
                 .setExecution(Execution.newBuilder().build())
                 .build();
   
         Execution executionCreated = executionsClient.createExecution(executionRequest);
   
         System.out.println("Created execution: " + executionCreated.getName());
   
         // Wait for execution to finish using exponential backoff
         long backoffDelay = 1_000;
         Execution execution;
         System.out.println("Poll for result...");
         while (true) {
           execution = executionsClient.getExecution(executionCreated.getName());
   
           // Check if execution finished
           if (execution.getState() != Execution.State.ACTIVE) {
             break;
           }
   
           // Wait using exponential backoff
           System.out.println("- Waiting for results...");
           Thread.sleep(backoffDelay);
           backoffDelay *= 2;
         }
   
         System.out.println("Execution finished with state: " + execution.getState().name());
         System.out.println("Execution results: " + execution.getResult());
       }
     }
   }

   Node.js

   const {ExecutionsClient} = require('@google-cloud/workflows');
   const client = new ExecutionsClient();
   /**
    * TODO(developer): Uncomment these variables before running the sample.
    */
   // const projectId = 'my-project';
   // const location = 'us-central1';
   // const workflow = 'myFirstWorkflow';
   // const searchTerm = '';
   
   /**
    * Executes a Workflow and waits for the results with exponential backoff.
    * @param {string} projectId The Google Cloud Project containing the workflow
    * @param {string} location The workflow location
    * @param {string} workflow The workflow name
    * @param {string} searchTerm Optional search term to pass to the Workflow as a runtime argument
    */
   async function executeWorkflow(projectId, location, workflow, searchTerm) {
     /**
      * Sleeps the process N number of milliseconds.
      * @param {Number} ms The number of milliseconds to sleep.
      */
     function sleep(ms) {
       return new Promise(resolve => {
         setTimeout(resolve, ms);
       });
     }
     const runtimeArgs = searchTerm ? {searchTerm: searchTerm} : {};
     // Execute workflow
     try {
       const createExecutionRes = await client.createExecution({
         parent: client.workflowPath(projectId, location, workflow),
         execution: {
           // Runtime arguments can be passed as a JSON string
           argument: JSON.stringify(runtimeArgs),
         },
       });
       const executionName = createExecutionRes[0].name;
       console.log(`Created execution: ${executionName}`);
   
       // Wait for execution to finish, then print results.
       let executionFinished = false;
       let backoffDelay = 1000; // Start wait with delay of 1,000 ms
       console.log('Poll every second for result...');
       while (!executionFinished) {
         const [execution] = await client.getExecution({
           name: executionName,
         });
         executionFinished = execution.state !== 'ACTIVE';
   
         // If we haven't seen the result yet, wait a second.
         if (!executionFinished) {
           console.log('- Waiting for results...');
           await sleep(backoffDelay);
           backoffDelay *= 2; // Double the delay to provide exponential backoff.
         } else {
           console.log(`Execution finished with state: ${execution.state}`);
           console.log(execution.result);
           return execution.result;
         }
       }
     } catch (e) {
       console.error(`Error executing workflow: ${e}`);
     }
   }
   
   executeWorkflow(projectId, location, workflowName, searchTerm).catch(err => {
     console.error(err.message);
     process.exitCode = 1;
   });
   

   Python

   import time
   
   from google.cloud import workflows_v1
   from google.cloud.workflows import executions_v1
   
   from google.cloud.workflows.executions_v1.types import executions
   
   # TODO(developer): Update and uncomment the following lines.
   # project_id = "YOUR_PROJECT_ID"
   # location = "YOUR_LOCATION"  # For example: us-central1
   # workflow_id = "YOUR_WORKFLOW_ID"  # For example: myFirstWorkflow
   
   # Initialize API clients.
   execution_client = executions_v1.ExecutionsClient()
   workflows_client = workflows_v1.WorkflowsClient()
   
   # Construct the fully qualified location path.
   parent = workflows_client.workflow_path(project_id, location, workflow_id)
   
   # Execute the workflow.
   response = execution_client.create_execution(request={"parent": parent})
   print(f"Created execution: {response.name}")
   
   # Wait for execution to finish, then print results.
   execution_finished = False
   backoff_delay = 1  # Start wait with delay of 1 second.
   print("Poll for result...")
   
   # Keep polling the state until the execution finishes,
   # using exponential backoff.
   while not execution_finished:
       execution = execution_client.get_execution(
           request={"name": response.name}
       )
       execution_finished = execution.state != executions.Execution.State.ACTIVE
   
       # If we haven't seen the result yet, keep waiting.
       if not execution_finished:
           print("- Waiting for results...")
           time.sleep(backoff_delay)
           # Double the delay to provide exponential backoff.
           backoff_delay *= 2
       else:
           print(f"Execution finished with state: {execution.state.name}")
           print(f"Execution results: {execution.result}")

Run the sample code

You can run the sample code and execute your workflow. Executing a workflow runs the deployed workflow definition associated with the workflow.

1. To run the sample, first install dependencies:

   C#

   dotnet restore

   Go

   go mod download

   Java

   mvn compile

   Node.js

   npm install -D tsx

   Python

   pip3 install -r requirements.txt

2. Run the script:

   C#

   GOOGLE_CLOUD_PROJECT=PROJECT_ID LOCATION=CLOUD_REGION WORKFLOW=WORKFLOW_NAME dotnet run

   Go

   GOOGLE_CLOUD_PROJECT=PROJECT_ID LOCATION=CLOUD_REGION WORKFLOW=WORKFLOW_NAME go run .

   Java

   GOOGLE_CLOUD_PROJECT=PROJECT_ID LOCATION=CLOUD_REGION WORKFLOW=WORKFLOW_NAME mvn compile exec:java -Dexec.mainClass=com.example.workflows.WorkflowsQuickstart

   Node.js

   npx tsx index.js

   Python

   GOOGLE_CLOUD_PROJECT=PROJECT_ID LOCATION=CLOUD_REGION WORKFLOW=WORKFLOW_NAME python3 main.py

   Replace the following:

   • PROJECT_ID: your Google Cloud project name
   • CLOUD_REGION: the location of your workflow (default: us-central1)
   • WORKFLOW_NAME: your workflow name (default: myFirstWorkflow)

   The output is similar to the following:

   Execution finished with state: SUCCEEDED
   Execution results: ["Thursday","Thursday Night Football","Thursday (band)","Thursday Island","Thursday (album)","Thursday Next","Thursday at the Square","Thursday's Child (David Bowie song)","Thursday Afternoon","Thursday (film)"]
   

Pass data in an execution request

Depending on the client library language, you can also pass a runtime argument in an execution request. For example:

public class ExecuteWorkflowWithArgumentsSample
{
    /// <summary>
    /// Execute a workflow with arguments and return the execution operation.
    /// </summary>
    /// <param name="projectID">Your Google Cloud Project ID.</param>
    /// <param name="locationID">The region where your workflow is located.</param>
    /// <param name="workflowID">Your Workflow ID.</param>
    /// <returns>
    /// An Execute object representing the completed workflow execution.
    /// </returns>
    public async Task<Execution> ExecuteWorkflowWithArguments(
        string projectId = "YOUR-PROJECT-ID",
        string locationID = "YOUR-LOCATION-ID",
        string workflowID = "YOUR-WORKFLOW-ID")
    {
        // Initialize the client.
        ExecutionsClient client = await ExecutionsClient.CreateAsync();

        // Build the parent location path.
        WorkflowName parent = new WorkflowName(projectId, locationID, workflowID);

        // Serialize the argument.
        string argument = JsonSerializer.Serialize(new
        {
            searchTerm = "Cloud"
        });

        // Create an execution request.
        CreateExecutionRequest createExecutionRequest = new CreateExecutionRequest
        {
            ParentAsWorkflowName = parent,
            Execution = new Execution
            {
                Argument = argument,
            }
        };

        // Execute the operation and recieve the execution.
        Execution execution = await client.CreateExecutionAsync(createExecutionRequest);
        Console.WriteLine("- Execution started...");

        TimeSpan backoffDelay = TimeSpan.FromSeconds(1);
        TimeSpan maxBackoffDelay = TimeSpan.FromSeconds(16);

        // Keep polling the state until the execution finishes, using exponential backoff.
        while (execution.State == Execution.Types.State.Active)
        {
            await Task.Delay(backoffDelay);

            // Implement exponential backoff by doubling the delay, but limiting it to a practical duration.
            backoffDelay = (backoffDelay < maxBackoffDelay) ? backoffDelay * 2 : maxBackoffDelay;

            execution = await client.GetExecutionAsync(execution.Name);
        }

        // Print results.
        Console.WriteLine($"Execution finished with state: {execution.State}");
        Console.WriteLine($"Execution results: {execution.Result}");

        // Return the fetched execution.
        return execution;
    }
}

import (
	"context"
	"encoding/json"
	"fmt"
	"io"
	"time"

	workflowexecutions "google.golang.org/api/workflowexecutions/v1"
)

// Execute a workflow with arguments and print the execution results.
//
// For more information about Workflows see:
// https://cloud.google.com/workflows/docs/overview
func executeWorkflowWithArguments(w io.Writer, projectID, workflowID, locationID string) error {
	// TODO(developer): Uncomment and update the following lines:
	// projectID := "YOUR_PROJECT_ID"
	// workflowID := "YOUR_WORKFLOW_ID"
	// locationID := "YOUR_LOCATION_ID"

	ctx := context.Background()

	// Construct the location path.
	parent := fmt.Sprintf("projects/%s/locations/%s/workflows/%s", projectID, locationID, workflowID)

	// Create execution client.
	client, err := workflowexecutions.NewService(ctx)
	if err != nil {
		return fmt.Errorf("workflowexecutions.NewService error: %w", err)
	}

	// Get execution service.
	service := client.Projects.Locations.Workflows.Executions

	// Create argument.
	argument := struct {
		SearchTerm string `json:"searchTerm"`
	}{
		SearchTerm: "Cloud",
	}

	// Encode argument to JSON.
	argumentEncoded, err := json.Marshal(argument)
	if err != nil {
		return fmt.Errorf("json.Marshal error: %w", err)
	}

	// Build and run the new workflow execution adding the argument.
	res, err := service.Create(parent, &workflowexecutions.Execution{
		Argument: string(argumentEncoded),
	}).Do()
	if err != nil {
		return fmt.Errorf("service.Create.Do error: %w", err)
	}
	fmt.Fprintln(w, "- Execution started...")

	// Set initial value for backoff delay in one second.
	backoffDelay := time.Second

	for res.State == "ACTIVE" {
		time.Sleep(backoffDelay)

		// Request the updated state for the execution.
		getReq := service.Get(res.Name)
		res, err = getReq.Do()
		if err != nil {
			return fmt.Errorf("getReq error: %w", err)
		}

		// Double the delay to provide exponential backoff (capped at 16 seconds).
		if backoffDelay < time.Second*16 {
			backoffDelay *= 2
		}
	}

	fmt.Fprintf(w, "Execution finished with state: %s\n", res.State)
	fmt.Fprintf(w, "Execution arguments: %s", res.Argument)
	fmt.Fprintf(w, "Execution results: %s\n", res.Result)

	return nil
}

public class ExecuteWithArguments {
  public static void main(String[] args)
      throws IOException, InterruptedException, ExecutionException {
    String projectId = "your-project-id";
    String location = "your-location"; // For example: us-central1
    String workflowId = "your-workflow-id";

    executeWorkflowWithArguments(projectId, location, workflowId);
  }

  public static void executeWorkflowWithArguments(
      String projectId, String location, String workflowId)
      throws IOException, InterruptedException, ExecutionException {
    try (ExecutionsClient executionsClient = ExecutionsClient.create()) {
      WorkflowName parent = WorkflowName.of(projectId, location, workflowId);

      CreateExecutionRequest executionRequest =
          CreateExecutionRequest.newBuilder()
              .setParent(parent.toString())
              .setExecution(
                  Execution.newBuilder().setArgument("{\"searchTerm\":\"Cloud\"}").build())
              .build();

      Execution executionCreated = executionsClient.createExecution(executionRequest);

      System.out.println("Created execution: " + executionCreated.getName());

      // Wait for execution to finish using exponential backoff
      long backoffDelay = 1_000;
      Execution execution;
      System.out.println("Poll for result...");
      while (true) {
        execution = executionsClient.getExecution(executionCreated.getName());

        // Check if execution finished
        if (execution.getState() != Execution.State.ACTIVE) {
          break;
        }

        // Wait using exponential backoff
        System.out.println("- Waiting for results...");
        Thread.sleep(backoffDelay);
        backoffDelay *= 2;
      }

      System.out.println("Execution finished with state: " + execution.getState().name());
      System.out.println("Execution results: " + execution.getResult());
    }
  }
}

// Execute workflow
try {
  const createExecutionRes = await client.createExecution({
    parent: client.workflowPath(projectId, location, workflow),
    execution: {
      argument: JSON.stringify({"searchTerm": "Friday"})
    }
});
const executionName = createExecutionRes[0].name;

import time

from google.cloud import workflows_v1
from google.cloud.workflows import executions_v1

from google.cloud.workflows.executions_v1.types import executions

# TODO(developer): Update and uncomment the following lines.
# project_id = "YOUR_PROJECT_ID"
# location = "YOUR_LOCATION"  # For example: us-central1
# workflow_id = "YOUR_WORKFLOW_ID"  # For example: myFirstWorkflow

# Initialize API clients.
execution_client = executions_v1.ExecutionsClient()
workflows_client = workflows_v1.WorkflowsClient()

# Construct the fully qualified location path.
parent = workflows_client.workflow_path(project_id, location, workflow_id)

# Execute the workflow adding an dictionary of arguments.
# Find more information about the Execution object here:
# https://cloud.google.com/python/docs/reference/workflows/latest/google.cloud.workflows.executions_v1.types.Execution
execution = executions_v1.Execution(
    name=parent,
    argument='{"searchTerm": "Cloud"}',
)

response = execution_client.create_execution(
    parent=parent,
    execution=execution,
)
print(f"Created execution: {response.name}")

# Wait for execution to finish, then print results.
execution_finished = False
backoff_delay = 1  # Start wait with delay of 1 second.
print("Poll for result...")

# Keep polling the state until the execution finishes,
# using exponential backoff.
while not execution_finished:
    execution = execution_client.get_execution(
        request={"name": response.name}
    )
    execution_finished = execution.state != executions.Execution.State.ACTIVE

    # If we haven't seen the result yet, keep waiting.
    if not execution_finished:
        print("- Waiting for results...")
        time.sleep(backoff_delay)
        # Double the delay to provide exponential backoff.
        backoff_delay *= 2
    else:
        print(f"Execution finished with state: {execution.state.name}")
        print(f"Execution results: {execution.result}")

For more information about passing runtime arguments, see Pass runtime arguments in an execution request.