Dapr workflow demos

Demos applications that use the Dapr Workflow building block.

Prerequisites

.NET 7 SDK
Dapr CLI
A REST client, such as cURL, or the VSCode REST client extension.

The VSCode REST client is configured as a recommended extension when opening this repo in VSCode.

Hello world workflow sample

The Hello World Workflow sample is a very basic workflow with just one activity that returns a random greeting. The workflow takes a name as input and returns a greeting with the name as output.

graph TD
    A[Start]
    B[CreateGreetingActivity]
    C[End]
    A -->|"input: {name}"| B -->|"output: Hi {name}"| C

Run the HelloWorldWorkflowSample app

Change to the HelloWorld directory and build the ASP.NET app:
```
cd HelloWorld
dotnet build
```
Run the app using the Dapr CLI:
```
dapr run --app-id hello-world --app-port 5065 --dapr-http-port 3500 dotnet run
```
Ensure the --app-port is the same as the port specified in the launchSettings.json file.
Start the HelloWorldWorkflow via the Workflow HTTP API using cURL, or use the helloworld.http file if you're using VSCode with the REST client:
```
curl -i -X POST http://localhost:3500/v1.0-alpha1/workflows/dapr/HelloWorldWorkflow/1234a/start \
  -H "Content-Type: application/json" \
  -d '{ "input" : "Marc"}'
```
Note that 1234a in the URL is the workflow instance ID. This can be any string you want.

Expected result:
```
{
  "instance_id": "<WORKFLOW_ID>"
}
```

Check the workflow status via Workflow HTTP API:

curl -i -X GET http://localhost:3500/v1.0-alpha1/workflows/dapr/HelloWorldWorkflow/1234a/status

Expected result:

{
    "WFInfo": {
        "instance_id": "<WORKFLOW_ID>"
    },
    "start_time": "2023-05-01T12:15:45Z",
    "metadata": {
        "dapr.workflow.custom_status": "",
        "dapr.workflow.input": "\"Marc"\",
        "dapr.workflow.last_updated": "2023-05-01T12:15:45Z",
        "dapr.workflow.name": "HelloWorldWorkflow",
        "dapr.workflow.output": "\"Hi Marc"\",
        "dapr.workflow.runtime_status": "COMPLETED"
    }
}

OrderService workflow sample

The OrderService Workflow sample is a workflow that processes an order. The workflow takes an order payload as input and returns an order result as output. The workflow uses these activities:

NotifyActivity: Notifies the customer of the progress of the order.
CheckInventoryActivity: Checks if the inventory is sufficient.
ProcessPaymentActivity: Processes the payment.
UpdateInventoryActivity: Updates the inventory after checking it again.
RefundPaymentActivity: Refunds the payment if the UpdateInventoryActivity throws an exception.

graph TD
    A[Start]
    B[NotifyActivity]
    C[CheckInventoryActivity]
    X{Sufficient Inventory?}
    D[ProcessPaymentActivity]
    E[UpdateInventoryActivity]
    F[RefundPaymentActivity]
    BB[NotifyActivity]
    BBB[NotifyActivity]
    XX{Sufficient Inventory?}
    Z[End]
    A --> |OrderPayload| B --> C
    C --> X
    X -->|Yes| D
    X -->|"No - OrderResult(Processed:false)"| Z
    D --> E --> XX
    XX -->|Yes| BB --> |"OrderResult(Processed:true)"| Z
    XX -->|No| BBB --> F --> |"OrderResult(Processed:false)"| Z

The CheckInventoryActivity and UpdateInventoryActivity classes use Dapr's state management building block to manage the inventory in a Redis state store.

Next to the workflow, this application has an InventoryController with the following endpoints:

GET http://localhost:5064/inventory: retrieves the inventory
DELETE http://localhost:5064/inventory: clears the inventory
POST http://localhost:5064/inventory/restock: restocks the inventory

The InventoryController also uses Dapr's state management building block.

Run the OrderServiceWorkflowSample app

Change to the OrderService directory and build the ASP.NET app:
```
cd OrderService
dotnet build
```
Run the app using the Dapr CLI:
```
dapr run --app-id order-processor --app-port 5064 --dapr-http-port 3500 --resources-path ./Resources dotnet run
```
Ensure the --app-port is the same as the port specified in the launchSettings.json file.
Check the inventory using cURL, or use the OrderService.http file if you're using VSCode with the REST client:
```
curl -X POST http://localhost:5064/inventory/restock
```
If the quantity of the items is not 0, clear the inventory by running:
```
curl -X POST http://localhost:5064/inventory/clear
```
Try ordering 100 paperclips while the inventory is not sufficient. Start the OrderProcessingWorkflow via the Workflow HTTP API:
```
curl -i -X POST http://localhost:3500/v1.0-alpha1/workflows/dapr/OrderProcessingWorkflow/1234a/start \
  -H "Content-Type: application/json" \
  -d '{ "input" : {"Name": "Paperclips", "Quantity": 100}}'
```
Note that 1234a in the URL is the workflow instance ID. This can be any string you want.

Expected result:
```
{
  "instance_id": "<WORKFLOW_ID>"
}
```
Pay attention to the console output. A message will appear that indicates the inventory is insufficient.

Check the workflow status via Workflow HTTP API:

curl -i -X GET http://localhost:3500/v1.0-alpha1/workflows/dapr/OrderProcessingWorkflow/1234a/status

Expected result:

{
  "WFInfo": {
    "instance_id": "<WORKFLOW_ID>"
  },
  "start_time": "2023-05-01T12:15:45Z",
  "metadata": {
    "dapr.workflow.custom_status": "\"Stopped order process due to insufficient inventory.\"",
    "dapr.workflow.input": "{\"Name\":\"Paperclips\",\"Quantity\":100}",
    "dapr.workflow.last_updated": "2023-05-01T12:15:45Z",
    "dapr.workflow.name": "OrderProcessingWorkflow",
    "dapr.workflow.output": "{\"Processed\":false}",
    "dapr.workflow.runtime_status": "COMPLETED"
  }
}

Depending on how quick the status is retrieved after starting the workflow, the dapr.workflow.runtime_status could still be "RUNNING". Repeat the GET status request until the status is "COMPLETED".

Restock the inventory:

curl -X POST http://localhost:5064/inventory/restock

Expected result: HTTP 200 OK

Try ordering paperclips again, now within the limits of the inventory. Start the OrderProcessingWorkflow via the Workflow HTTP API:
```
curl -i -X POST http://localhost:3500/v1.0-alpha1/workflows/dapr/OrderProcessingWorkflow/1234b/start \
 -H "Content-Type: application/json" \
 -d '{ "input" : {"Name": "Paperclips", "Quantity": 100}}'
```
Note that 1234b in the URL is the workflow instance ID. This can be any string you want.

Expected result:
```
{
  "instance_id": "<WORKFLOW_ID>"
}
```
Pay attention to the console output. Messages will appear that indicate the inventory is sufficient and payment has been processed successfully.

Check the workflow status via Workflow HTTP API:

curl -i -X GET http://localhost:3500/v1.0-alpha1/workflows/dapr/OrderProcessingWorkflow/1234b/status

Expected result:

{
  "WFInfo": {
    "instance_id": "<WORKFLOW_ID>"
  },
  "start_time": "2023-05-01T12:22:25Z",
  "metadata": {
    "dapr.workflow.custom_status": "",
    "dapr.workflow.input": "{\"Name\":\"Paperclips\",\"Quantity\":100}",
    "dapr.workflow.last_updated": "2023-05-01T12:22:37Z",
    "dapr.workflow.name": "OrderProcessingWorkflow",
    "dapr.workflow.output": "{\"Processed\":true}",
    "dapr.workflow.runtime_status": "COMPLETED"
  }
}

Inspect the logs in ZipKin: localhost:9411/zipkin. Find the entry marked order-processor:create_orchestration||orderprocessingworkflow and show the details. You'll now see a timeline of the workflow at the top, and the activities underneath.

Resources

More information

Any questions or comments about this sample? Join the Dapr discord and post a message the #workflow channel. Have you made something with Dapr? Post a message in the #show-and-tell channel, we love to see your creations!

kendallroden / dapr-workflow-demos Goto Github PK