Skip to main content

Running Your First Workflow


Now that you've created your first WfSpec, let's explore the different ways to run it. LittleHorse provides multiple methods to execute workflows, each suited for different use cases.

note

This tutorial assumes you have completed the Your First WfSpec guide.

Running with lhctl​

The simplest way to run a workflow is using the command-line interface. The lhctl command provides a quick way to execute and monitor workflows:

# Run the workflow with a name parameter
lhctl run getting-started name Obi-Wan

You'll see output like:

{
  "id":  {
    "id":  "69881bf9aaf84c2f84bdde7ceb2dd105"
  },
  "wfSpecId":  {
    "name":  "getting-started",
    "majorVersion":  0,
    "revision":  0
  },
  "oldWfSpecVersions":  [],
  "status":  "RUNNING",
  "greatestThreadrunNumber":  0,
  "startTime":  "2024-12-17T01:53:47.212Z",
  "threadRuns":  [
    {
      "wfSpecId":  {
        "name":  "getting-started",
        "majorVersion":  0,
        "revision":  0
      },
      "number":  0,
      "status":  "RUNNING",
      "threadSpecName":  "entrypoint",
      "startTime":  "2024-12-17T01:53:47.296Z",
      "childThreadIds":  [],
      "haltReasons":  [],
      "currentNodePosition":  1,
      "handledFailedChildren":  [],
      "type":  "ENTRYPOINT"
    }
  ],
  "pendingInterrupts":  [],
  "pendingFailures":  []
}
info

The lhctl command is perfect for development and testing scenarios. It's also commonly used in CI/CD pipelines for automation.

Running from the Dashboard​

LittleHorse provides a visual interface for running and monitoring workflows:

  1. Navigate to http://localhost:8080 in your browser
  2. Click on the "getting-started" workflow
  3. Click "Execute"
  4. Enter "Obi-Wan" in the name field
  5. Click "Execute Workflow"

Dashboard Execute Workflow

The dashboard will show you the workflow's execution status in real-time, making it ideal for debugging.

Running with the Java SDK​

For production applications, you'll typically want to run workflows programmatically. Create a new file called RunWorkflow.java:

package io.littlehorse.tutorial;

import io.littlehorse.sdk.common.config.LHConfig;
import io.littlehorse.sdk.client.LHClient;

public class RunWorkflow {
    public static void main(String[] args) throws Exception {
        // Initialize the client
        LHConfig config = new LHConfig();
        LHClient client = new LHClient(config);

        // Run the workflow
        String wfRunId = client.runWorkflow(
            "getting-started",
            java.util.Map.of("name", "Anakin"),
            null
        ).getId();

        System.out.println("Started workflow with ID: " + wfRunId);
    }
}
note

The Java SDK provides the most flexibility for running workflows. You can integrate it into your existing applications and handle workflow execution programmatically.

Choosing the Right Method​

Each method has its ideal use case:

MethodBest For
lhctlDevelopment, testing, and CI/CD pipelines
DashboardDebugging, monitoring, and team collaboration
Java SDKProduction applications and system integration

Remember, all methods interact with the same LittleHorse server, so you can mix and match them based on your needs.

Wrapping Up​

You've learned how to run workflows using different methods and monitor their execution. In the next lesson, we'll dive deeper into debugging workflows and handling errors.

If you haven't already: