n8n Error Handling Tutorial: Retries, Continue on Fail, and Error Notifications

The biggest risk in automation is failing silently: an API is temporarily unavailable, data is in the wrong format, credentials expire… If no one notices, notifications may go out unsent for days. This section introduces n8n's error-handling features.

1. Retry automatically (Retry On Fail)

Many errors are temporary (an unstable network, an occasional API timeout). In the node's Settings tab:

• Retry On Fail: Turn it on
• Max Tries: The maximum number of attempts, such as 3
• Wait Between Tries (ms): The interval between attempts, such as 5000 (5 seconds)

This is useful for nodes that call external APIs (HTTP Request and nodes for various services).

2. What to do when a node fails (On Error)

Also in the node's Settings tab, On Error has three options:

Use an error output branch

When processing multiple items, one failure shouldn't stop the others:

                    ┌─ success ─▶ [Mark as processed]
[Call API for each item] ─────┤
                    └─ error ───▶ [Log to the "Failed items" spreadsheet]

The failed item includes an error message, which you can read with {{ $json.error.message }}.

3. Stop proactively: Stop and Error

When data doesn't meet expectations, stop proactively instead of letting later nodes produce incorrect results:

[Get order] ─▶ [If amount is empty?] ─true─▶ [Stop and Error: "Order {{ $json.id }} is missing an amount"]
                                      └false─▶ [Continue processing]

The Stop and Error node causes the execution to fail and triggers the error workflow described below.

4. Global error notifications: Error Trigger

The most important step: create a dedicated workflow to handle errors and automatically notify you whenever any workflow fails.

Create an error workflow

1. Add a workflow and name it "Error notifications".
2. Select Error Trigger as the first node.
3. Connect a notification node (Telegram, Email, Slack, etc.) and use this message:

⚠️ n8n workflow execution failed
Workflow: {{ $json.workflow.name }}
Node: {{ $json.execution.lastNodeExecuted }}
Error: {{ $json.execution.error.message }}
Time: {{ $now.toFormat('yyyy-MM-dd HH:mm') }}
View: {{ $json.execution.url }}

4. Save it. A workflow that starts with Error Trigger works without being published.

Assign it to other workflows

For each workflow you want to monitor, go to Menu → Settings → Error workflow and select "Error notifications."

After that, a notification is sent automatically whenever this workflow ends with a "Failed" status.

The error workflow is triggered only when an automated execution (such as a schedule or webhook) fails. A failed manual test does not trigger it.

5. View and rerun executions

• Filter by Error status in Executions to see all failed executions.
• Open a failed execution and click Debug in editor to test again with the data from that run.
• After fixing the problem, you can rerun the failed execution.

6. Make data traceable

• In Settings, configure how execution records are saved. You can save only failed executions to reduce database usage.
• For important workflows, write a record to a spreadsheet or database at the end so you can later look up which item was processed and when.

Recommended practices

Make it a habit to configure error notifications as soon as you create a workflow that needs to run long term.

Updated October 12, 2026