n8n

n8n Error Workflow: what it catches, and the 5 failures it will never see

The n8n error workflow is essential, but it only reacts to failed executions. Here are the five failures that slip past it and how to cover them.

By Abdellah El Kharraze · · 2 min read

Key takeaways
  • The Error Workflow fires when a production execution fails.
  • It does not see: a mute trigger, a stopped instance, a deactivated workflow, a drop in volume, slowness.
  • It has to be attached to each workflow, one by one.
  • Complement it with external monitoring.

The Error Workflow is n8n's native alerting mechanism: a workflow that starts with an Error Trigger node and that you designate, in another workflow's settings, as the one to run on failure. It receives the workflow name, the execution concerned and the error message. You do what you like with it: an email, a Slack or Telegram message.

What it does well

It warns you immediately when an execution crashes, with the technical detail: which node, which message. For a third-party API error or malformed data, that is exactly what you need. Set it up on all your production workflows.

The five failures it will never see

  1. The mute trigger. If the webhook receives no more calls or the scheduled trigger no longer fires, there is no execution, so no failure, so no alert.
  2. The stopped instance. An n8n that is off cannot run the error workflow. The most serious failure is the quietest.
  3. The deactivated workflow. A colleague switches it off for a test and forgets to switch it back on. No error.
  4. The drop in volume. The workflow runs but handles ten times fewer items than before. Every execution succeeds.
  5. Slowness. An execution that goes from ten seconds to ten minutes still succeeds.

Two configuration traps

The error workflow is set per workflow. A new workflow created without that setting alerts nobody. And if the error workflow itself fails, for example because your messaging token has expired, you will not know either.

How to cover the rest

You need an observer placed outside the instance, which regularly reads the state of workflows and their executions through the API, and which knows each one's normal rhythm. That is what the guide Monitoring n8n in production describes, and what Nokto does, including an audit of workflows without error handling.

Frequently asked questions

Should I keep the Error Workflow if I use a monitoring tool?

Yes. It gives the technical detail of a failure the moment it happens. External monitoring covers what it cannot see.

AE
Abdellah El Kharraze

Founder of Nokto and Owepo, based in Nice, France. Network and security engineer for large enterprises until 2016, then ten years building n8n and Make automations for small businesses.

Your first monitored workflow in ten minutes.

Connect an n8n instance or a Make team: Nokto discovers your workflows, learns their rhythm and warns you at the first anomaly.

14 days free · no credit card · cancel anytime