Execution History & Failed Runs
Every integration workflow execution is recorded. This guide explains how to read execution history, debug failures, and understand execution statuses.
Viewing execution history
- Go to Integrations → Execution History
- You'll see a table of all executions across all your workflows
- Filter by workflow, status, or date range
Each row shows:
- Workflow name
- Trigger type — e.g.
order.created - Status — RUNNING, COMPLETED, or FAILED
- n8n Execution ID — links to the raw n8n execution
- Started at / Completed at
Execution statuses
| Status | Meaning |
|---|---|
| RUNNING | n8n is currently executing the workflow |
| COMPLETED | All actions ran successfully |
| FAILED | n8n reported an error; one or more actions did not complete |
RUNNING executions that are more than 5 minutes old usually indicate a network issue between LaunchStore and n8n. Check the n8n logs if this happens.
Understanding a FAILED execution
When an execution fails:
- The
integration_executionstable recordsstatus = FAILED - The n8n execution ID is stored — you can look up the full error in n8n's execution history
- The dashboard shows the failure with a red indicator
Common failure reasons
| Error | Likely cause |
|---|---|
401 Unauthorized | API key expired or revoked — reconnect the integration |
404 Not Found | The resource doesn't exist in the target system (e.g. HubSpot contact ID mismatch) |
422 Unprocessable Entity | A required field is missing or has an invalid format |
429 Too Many Requests | Rate limit hit — n8n will retry automatically on most integrations |
Network timeout | Temporary connectivity issue between n8n and the external API |
Reconnecting a broken integration
If your API key has been rotated or revoked:
- Go to Settings → Integrations
- Find the affected integration
- Click Edit → update the API key → Save & Test
- Re-activate any workflows that were using it
Real-time WebSocket updates
The Execution History page receives live updates via WebSocket. You don't need to refresh — new executions appear as they start and complete. Status changes from RUNNING to COMPLETED or FAILED update in place.
Execution payload (context)
Each execution record stores the trigger event payload in the context column. Click on any execution to view:
- Trigger payload — the full Medusa event data (order, customer, fulfillment)
- n8n execution ID — click to view the raw n8n execution log
This makes it easy to debug field mapping issues — you can see exactly what data was sent to the workflow.
Backpressure and dropped events
LaunchStore enforces per-store limits:
- 100 maximum concurrent executions
- 1,000 maximum queued executions
If your store receives a spike of events that exceeds these limits (e.g. a flash sale with 5,000 orders in 1 minute), events beyond the limit are dropped with a warning log rather than queued indefinitely.
For high-volume stores, consider:
- Increasing concurrency in the store settings (Pro/Enterprise plans)
- Using n8n's built-in rate limiting per workflow
- Batching order processing in off-peak hours
Workflow execution analytics
The Workflow Analytics dashboard (under Integrations → Analytics) shows:
- Total executions over time
- Success rate per workflow
- Average execution duration
- Most frequently triggered workflows
Use this to identify workflows that are failing frequently or running slower than expected.