- An uptime probe checks that an address responds.
- An n8n that responds can have every workflow stopped.
- Workflow monitoring looks at executions, not the server.
- The two complement each other on a self-hosted instance.
An uptime probe calls an address at regular intervals and alerts if it does not answer. It is simple, cheap, and essential for a server. But an n8n instance that responds is not an instance whose workflows are working.
What an uptime probe sees
It knows whether the server is reachable, whether the certificate is valid and how long the page takes to answer. It knows nothing about what happens inside.
What it does not see
- A workflow that fails on every execution.
- A deactivated workflow.
- A trigger that no longer fires.
- A stuck execution queue while the interface still responds.
- Anything running on Make, where there is no server of yours to probe.
The two approaches side by side
| Question | Uptime probe | Workflow monitoring |
|---|---|---|
| Is the server responding? | Yes | Yes (unreachable instance) |
| Did this workflow fail? | No | Yes |
| Is this workflow still running? | No | Yes |
| Is it handling the usual volume? | No | Yes |
| Works for Make? | No | Yes |
Do you need both?
On self-hosted n8n, yes: the uptime probe tells you within seconds that a server is down, and workflow monitoring covers everything else. On n8n Cloud and Make, workflow monitoring is enough, since you do not operate the server.
Nokto also reports an unreachable instance, because it queries its API every two minutes. See how it works.
Frequently asked questions
Can an uptime probe monitor an n8n webhook?
It can check that the address responds, not that the workflow behind it correctly handles real calls.