Running Flows

Running a flow means executing its sequence of actions, triggers, conditions, and transformations. How a flow runs depends on its structure and configuration. This section explains the different ways you can launch flows in FlowRunner™.

Basic Rules for Launching Flows¶

To ensure your flows start smoothly, follow these basic rules:

Enable the Flow: A flow must be in the LIVE state to run:
Starting Without a Schedule: If a flow doesn’t have a schedule, it must be started with a special command called CallFlow.

Note

There is an exception to this rule for flows that start with a trigger. In this case, the CallFlow command is optional. The flow can begin with the activation of the first trigger. However, initiating the flow with the CallFlow command can be beneficial. See the section on the CallFlow API for more details.
Scheduled Flows: When a flow has a schedule, it will automatically run according to that schedule. The CallFlow command is not needed as the scheduler handles the execution.
Multiple Executions: A single flow can have multiple executions, some running in parallel.
Unique Identifiers: Every execution of a flow has a unique identifier called executionId.

Below is a diagram illustrating the "launch rules":

flowchart TD
    GOLIVE(Flow is LIVE) --> STARTS_WITH_ACTION
    STARTS_WITH_ACTION{Does the Flow start with an Action?} -->|YES|HASSCHEDULE 
    STARTS_WITH_ACTION --> |No, starts with a Trigger|START_TRIGGER
    HASSCHEDULE{Has a Schedule?} -->|YES|FLOW_STARTS
    HASSCHEDULE -->|NO|NEEDS_CALL_FLOW
    NEEDS_CALL_FLOW --> FLOW_STARTS
    START_TRIGGER --> FLOW_STARTS
    NEEDS_CALL_FLOW("`Use *CallFlow*`")
    START_TRIGGER("`Activate Trigger.
    ...
    May optionally 
    use *CallFlow*`")
    FLOW_STARTS(Flow Instance is Started)

The sections below provide more detail on each scenario.

Flow without a Schedule¶

For a flow without a schedule, once you enable it (putting it in the LIVE state), it won’t run automatically. If the flow starts with a trigger, the trigger needs to be activated for the flow to run. If the flow starts with an action, it must be started using the CallFlow command/API.

Flow with a Schedule¶

For a scheduled flow that is in the LIVE state, the scheduler starts new executions automatically. The CallFlow command/API is not needed in this case. If the flow starts with an action, the action block will execute with each new iteration of the schedule. If the flow starts with a trigger, each new execution will wait for the trigger to be activated.

CallFlow Command/API¶

The CallFlow command creates a new execution of a flow. This command identifies the flow to execute and can optionally accept a key/value structure (object) to pass into the executed flow. The CallFlow command is available in several formats:

FlowRunner™ Action: Execute a flow from another flow.
Codeless block: Execute a flow from your UI application created with UI Builder or from Backendless Cloud Code. You can find the Call Flow Codeless block in the Codeless Logic Editor. Every flow you create will have a dedicated menu item in the FLOWRUNNER section: The Call Flow Codeless block returns an object which contains the executionId property. The value of the property can be used later on to activate a trigger in the same flow execution. Below is an example of how to obtain the executionId value from the Call Flow response:
REST API endpoint: Execute a flow from any third-party system.

Endpoint:
```
POST https://xxxx.backendless.app/api/automation/flow/activate-by-name
```
HTTP Headers:
```
Content-Type: application/json
user-token: value 
```
The user-token header is optional. When it is used, the activated flow receives the identity of the user represented by the token.

Request Body:
```
{
  "name": "Flow name. Required Value",
  "initialData": {
    "A valid JSON object containing initial flow data. Optional"
  }
}
```
The initialData property in the request body is optional. When provided, the value is passed into the flow execution. It will be available through the Expression Editor interface.

Response Body
```
{
  "executionId": "String"
}
```

The return value from the CallFlow operation is an executionId assigned to the created flow execution. This ID is important if you need to "target" triggers from that specific execution. When calling a trigger, you can pass the executionId parameter to activate the trigger in a specific execution.