Why a successful test is insufficient
An original report describes this situation with self-hosted n8n, a public tunnel and an Instagram Login integration. It does not establish a universal fix. Q1 n8n distinguishes test and production webhooks; an editor test listener does not replace the published production endpoint. Q2
A request you send yourself also checks only the route you used. It does not prove that the service attempted to deliver a real messaging event for the intended account. Avoid combining instructions for different Instagram integration variants without establishing which one your application uses.
Locate the missing stage
- Define a specific event. Send a permitted test message between accounts appropriate for your setup. Record the timestamp and timezone, application identity, receiving account and expected event type. Use a neutral test message instead of personal conversation content.
- Compare production addresses. Compare the callback stored in the application with the Production URL of the published workflow, including path and expected HTTP method. Successful callback verification is not the same operation as subsequent event delivery.
- Inspect the service-side evidence. Where the relevant interface provides them, examine event subscriptions, connected account, app mode and delivery information. If no delivery attempt exists, investigate the prerequisites of that specific API variant first. Do not copy permission names from another integration model.
- Correlate ingress logs. Search reverse-proxy or tunnel logs for the timestamp and path. If a request arrives, record its method, response status and forwarding destination. If it does not, a processing node behind the n8n webhook is not yet the first place to change.
- Account for n8n visibility. Check workflow identity, published version and execution-list settings. A missing visible execution alone does not prove that no request reached the public endpoint.
Complete a controlled end-to-end check
A real test should be traceable from the event to a harmless intended result. Keep correlation data and prevent automatic replies during investigation. Do not remove signature or access controls merely to obtain an apparent success. After correcting the identified stage, repeat one controlled event and compare all three observation points again.
Frequently asked questions
Should I simply switch the app to live mode?
This symptom does not establish that requirement. Account type, API variant, approved access and test setup must fit together. Check current requirements for the exact configuration you use.
What makes a useful handover?
Provide an anonymised connection diagram, n8n version, callback path, application and account type, and a timestamp with evidence from the three stages. Access tokens and complete private message histories are unnecessary.
Help with this case
We can help isolate the affected process and prepare a controlled next step. Discuss case A0158 with KAN KAI WORKS.
Prepare a findings note
Summarise your observations for your enquiry. The note is created in your browser and is not saved or sent here. You can then copy it into the contact form.
Unknown is fine. Do not enter passwords, customer data or complete confidential logs.
Support with this issue
We investigate n8n using your error message, versions and a narrowly scoped example.
Sources
Sources checked on 5 November 2026. Individual reports do not establish general prevalence.
