See llms.txt for all machine-readable content.

Back to Templates

Monitor self-hosted n8n uptime and workflow failures with Vivere

Created by

Created by: Vivere || vivere
Vivere

Last update

Last update 6 hours ago

Categories

Quick overview

Get alerted when your self-hosted n8n instance goes down, stops running schedules, or workflows fail. Every 5 minutes this workflow checks the n8n API for failed executions and pings a Vivere heartbeat monitor. If the instance stops, the pings stop, and Vivere alerts you from outside n8n.

How it works

  1. Runs every 5 minutes on a schedule.
  2. Reads up to 100 recent n8n executions with status "error" using the n8n API.
  3. Filters failures to only those within the configured look-back window and ignores failures from this watchdog workflow itself, or captures an API error if the n8n API call fails.
  4. Sends a success ping to your Vivere heartbeat monitor when no recent failures are found.
  5. When failures are found, fetches workflow names from the n8n API and builds a short text summary mapping workflow IDs to names.
  6. Sends a fail ping to Vivere containing the failure summary, or sends a fail ping describing the API error when the n8n API is unreachable or misconfigured.

Setup

  1. Create an n8n API key (Settings → n8n API) and add an n8n API credential with your instance base URL ending in /api/v1.
  2. Sign up at vivere.dev, create a Heartbeat monitor with a 5-minute period and a 5-minute grace, and copy its ping URL.
  3. Paste the Vivere ping URL into the workflow configuration and adjust lookbackMinutes (and the schedule interval) to match your desired check frequency.

Requirements

  • A self-hosted n8n instance with the public API enabled, and a free Vivere account (25 monitors, email and webhook alerts, 5-minute minimum period).

Customization

  • Check less often by changing the trigger, lookbackMinutes and the monitor's period together, or ignore noisy workflows by filtering their ids out in Check executions.

Additional info

Why check from outside: an error workflow or an in-n8n watchdog runs inside the instance it watches, so it can't alert you when that instance is down, a restart didn't come back, or the scheduler stalled. Here the missing ping is the alert. More on catching n8n workflows that stopped silently: https://vivere.dev/answers/n8n-workflows-stopped-no-alert?utm_source=n8n-template&utm_campaign=22556