Monitoring Workflows¶
After executing a workflow — whether manually or via a trigger — you can monitor its progress, inspect results, and debug failures through the FlowDrop admin UI.
Pipelines: Workflow Runs¶
Every workflow execution creates a Pipeline — a record of one complete run. Each node that executes within that run creates a Job.
Viewing Pipelines¶
Navigate to Administration > FlowDrop > Pipelines (/admin/flowdrop/pipelines).
This page shows all pipeline runs, most recent first. Each row displays:
| Column | Description |
|---|---|
| ID | Unique pipeline identifier |
| Workflow | The workflow that was executed |
| Status | Current pipeline status (see below) |
| Created | When the execution started |
Pipeline Statuses¶
| Status | Meaning |
|---|---|
| Pending | Pipeline is queued and waiting to be processed |
| Running | Pipeline is currently executing |
| Completed | All nodes finished successfully |
| Failed | One or more nodes failed during execution |
| Paused | Pipeline is paused (waiting for user input via an interrupt) |
| Cancelled | Pipeline was manually cancelled |
Jobs: Node Executions¶
Click on a pipeline to see its individual jobs. Each job represents one node's execution.
Job Statuses¶
| Status | Meaning |
|---|---|
| Pending | Node is waiting to execute (dependencies not yet resolved) |
| Running | Node is currently executing |
| Completed | Node finished successfully |
| Failed | Node encountered an error |
| Interrupted | Node is paused, waiting for user input |
| Skipped | Node was not executed (e.g., on an inactive branch of a gateway) |
Inspecting Job Details¶
Click on an individual job to see:
- Input data — What data the node received through its input ports
- Output data — What the node produced
- Error message — If the job failed, the error details
- Timing — When the job started and finished
This is the most useful view for debugging — you can trace exactly what data each node received and produced.
Debugging Common Issues¶
A node shows "Failed"¶
- Click on the failed job to see the error message.
- Common causes:
- Missing input — A required input port has no connection and no default value.
- External service error — For HttpRequest or AI nodes, the external service returned an error.
- Entity not found — An EntityQuery or EntityContext node could not find the expected entity.
- Type mismatch — The data arriving at an input port does not match the expected type.
Workflow completed but output is wrong¶
- Open the pipeline detail page.
- Click through each job in execution order.
- Compare the output data of each node with what you expected.
- The first node whose output is unexpected is where the problem lies.
Workflow is stuck in "Pending"¶
This usually means the asynchronous queue is not being processed. See Execution Modes — Queue Processing.
Workflow is stuck in "Paused"¶
The workflow hit an interrupt node (confirmation, choice, text input) and is waiting for a response. Either:
- Respond to the interrupt via the Playground interface
- Cancel the pipeline if the interrupt is no longer needed
Checking Logs¶
FlowDrop writes to Drupal's standard logging system.
Via the Admin UI¶
Navigate to Reports > Recent log messages (/admin/reports/dblog) and filter by type flowdrop.
Via Drush¶
# All FlowDrop logs
drush watchdog:show --type=flowdrop
# Trigger-specific logs
drush watchdog:show --type=flowdrop_trigger
# Recent errors only
drush watchdog:show --severity=error
Snapshot Cleanup¶
Over time, workflow execution snapshots accumulate. To manage storage:
- Navigate to Administration > FlowDrop > Snapshots > Cleanup (
/admin/flowdrop/snapshots/cleanup). - Or schedule periodic cleanup as part of your site maintenance routine.
Next Steps¶
- Execution Modes — Understand orchestrators and queue processing
- Troubleshooting — More detailed issue resolution