n8n Error Handling: How to Keep Your Workflows Running
An automation that fails quietly is worse than no automation at all. An invoice is never sent, a lead never reaches your CRM, and nobody notices until a customer complains. n8n has several built-in ways to catch errors, retry and alert you. This guide explains each one in plain language and how to combine them so your workflows keep running.
This post may contain affiliate links. If you buy through them, we may earn a commission at no extra cost to you.
Why workflows fail
Most failures come from a short list of causes:
An external service is slow or briefly down, or rate-limits your requests.
A credential or token expires.
Incoming data is missing a field or arrives in an unexpected format.
A service changes its API or the structure of its response.
The first cause is usually temporary and fixes itself with a retry. The others need a person to look, which is why alerts matter. If you are new to n8n, start with n8n for Beginners.
1. Turn on Retry On Fail for flaky steps
Every node has a Settings tab. Turn on Retry On Fail and choose how many tries to make and how long to wait between them. This is ideal for steps that call outside services, such as an HTTP request or a CRM update. A short wait of a few seconds handles most temporary hiccups without anyone noticing.
2. Decide what happens when a node fails
In the same Settings tab, the On Error option controls what the workflow does when a node still fails:
Stop Workflow: the default. The run ends and is marked as failed.
Continue: the workflow keeps going as if nothing happened. Use it only when the step is optional.
Continue (using error output): the node gets a second output for failed items, so you can route them to a different path, for example a spreadsheet of items to check later.
The error output is the most useful option for batches: one bad record no longer stops the other ninety-nine.
3. Create one error workflow for everything
An error workflow is a separate workflow that runs whenever another workflow fails. To set one up:
Create a new workflow that starts with the Error Trigger node.
Add the steps that should happen on failure, such as sending a message.
Open each workflow you want to protect, go to its Settings and choose your new workflow under Error Workflow.
The Error Trigger receives useful details: the name of the workflow, the node that failed, the error message, a link to the failed execution and when it happened. One error workflow can serve all your workflows.
One detail that confuses many people: the Error Trigger only fires for automatic, production runs. If you click to test a workflow manually and it fails, the error workflow will not run.
4. Send alerts where you will see them
Inside the error workflow, send the details to the channel you actually check: email, Slack, Telegram or a team chat. Include the workflow name, the failed node, the error message and the execution link, so you can open the problem in one click. Keep alerts short; a wall of text is easy to ignore.
5. Validate data before it causes trouble
Many errors start with bad input. Add an IF or Filter node early in the workflow to check that required fields exist, such as an email address on a new lead. Send incomplete items to a separate path instead of letting them break a later step. When the problem is serious, the Stop And Error node lets you end the run on purpose with a clear message, which then reaches your error workflow.
6. Review executions every week
The Executions list shows every run, successful or not. A quick weekly look shows patterns that alerts miss, such as a step that fails every Monday or a workflow that has quietly stopped running. Check how long your plan keeps execution history; on self-hosted setups you control it yourself. See n8n Cloud vs Self-Hosted: Costs and Trade-offs for the differences.
Where to start
You do not need all six on day one. Start with the workflows that touch money or customers, such as automating invoices with n8n and Google Sheets or sending website leads to your CRM. Add Retry On Fail to their external steps and connect them to one error workflow with an alert. Then extend the same setup to everything else, including scheduled jobs such as automated social media posting.
The bottom line
Good n8n error handling is a few simple habits: retry flaky steps, decide what each node should do on failure, route every failure to one error workflow, send a clear alert and check your executions weekly. An hour of setup saves you from the silent failures that cost customers and money.
Get one ready-to-use automation every week: subscribe to the Blog Spotlight Weekly newsletter.
Comments