Immediate failure when clicking "execute workflow"

[psoliveira80]

Describe the problem/error/question

Hi everyone. I’ve been up three nights researching this issue, so please help me. I’ve already reset the server and reinstalled N8N several times. This is a clean install. It only includes the Whatsapp Trigger node, and I entered the credentials (connection successful). When I click “Execute Workflow,” a popup immediately appears with this error. Does anyone have any idea what this could be?

1. I registered my domain with Cloudflare without a proxy.
2. I can’t access N8N directly via the IP address in the browser, only via the domain (could this be the problem?)
3. I have Traefik, Postgres, and Portainer installed in Docker.

What is the error message (if any)?

(#2200) Callback verification failed with the following errors: HTTP Status Code = 404; HTTP Message = Not Found

Please share your workflow

Share the output returned by the last node

{
    "code": 0,
    "message": "Bad request - please check your parameters",
    "level": "warning",
    "tags": {},
    "description": "(#2200) Callback verification failed with the following errors: HTTP Status Code = 404; HTTP Message = Not Found",
    "timestamp": 1753896557353,
    "context": {
        "data": {
            "error": {
                "message": "(#2200) Callback verification failed with the following errors: HTTP Status Code = 404; HTTP Message = Not Found",
                "type": "OAuthException",
                "code": 2200,
                "fbtrace_id": "A6klkf5dS8Cb8rW4JsaM9iI"
            }
        }
    },
    "functionality": "regular",
    "name": "NodeApiError",
    "node": {
        "parameters": {
            "whatsAppNotice": "",
            "updates": [
                "messages"
            ],
            "options": {}
        },
        "type": "n8n-nodes-base.whatsAppTrigger",
        "typeVersion": 1,
        "position": [
            -240,
            -96
        ],
        "id": "db5b7c97-c865-465e-bbae-816a34321326",
        "name": "WhatsApp Trigger",
        "webhookId": "9181244f-d371-4638-8081-208016f8997a",
        "credentials": {
            "whatsAppTriggerApi": {
                "id": "V6sEcVvoEAAW8VRn",
                "name": "WhatsApp OAuth account"
            }
        }
    },
    "messages": [
        "Request failed with status code 400"
    ],
    "httpCode": "400"
}

Information on your n8n setup

• n8n version: 1.104.1
• Database (default: SQLite): Postgres
• n8n EXECUTIONS_PROCESS setting (default: own, main):
• Running n8n via (Docker, npm, n8n cloud, desktop app): Docker Swarm
• Operating system: Ubuntu

[jabbson]

Could you please share how your webhook looks like for this trigger?
You can get it by clicking on this text (Webhook URLs):

[psoliveira80]

Hello, here is the trigger screen

[psoliveira80]

This only happens in the WhatsApp trigger, in a normal webhook trigger, it works normally.

{"data":{"waitingForWebhook":true}}

[psoliveira80]

Video Error Google Drive

I’ve already set up and taken down the server 12 times. 4 sleepless nights. Ghetto no more lol

@Samrat_Tamrakar do you found solution?

[jcuypers]

Hi, the error is regarding an oauth / authentication issue, its not about being able to use the webhook itself. Please verify your credentials and scopes etc etc with WhatsApp/n8n. Clear and restart of needed. Follow a guide from somewhere to be sure.

Also: here’s some info from a quick Google
https://www.reddit.com/r/n8n/comments/1k5jiha/whatsapp_trigger_failing_on_selfhosted_n8n_where/

And just to confirm: you can access your N8N instance via DNS from anywhere ?

Reg,
J.

[psoliveira80]

Yes, its avaiable: https://n8n.rpmincorporacoes.com.br/

Thanks for your reply. My problem is identical to his. If I enter the production URL on the meta page, it validates. My problem is this error on the test screen. From what I read in the thread you linked, his friend’s problem “solved itself,” but he doesn’t quite know how he solved it.

[jcuypers]

Hi, from what I understand from the link is that it only works when the workflow is active (in order for WhatsApp to validate whatever it needs to validate ). Test workflow will not work

Some “out of the box” idea that I can think of (no guarantees but it might work). I’ll describe the idea you need to see how you could implement it.

Let’s assume that your N8N workflow webhooks would be Whatsapp-WH-prod and whatsapp-WH-test. If you have bought a DNS name and control the zone you could create an alias like whatsapp-WH. You would configure this one as a callback URL in WhatsApp. Initially the whatsapp-WH name would point to prod to validate and after that you can change it to the test environment (make sure it has the same secrets etc). Also be aware that DNS changes might take time, since they are cached.

Reg
J.

[jcuypers]

Well.. I forgot it’s not an actual domain but a path by default. Webhook and webhook-test… But still I think the idea is still valid. You might be able to do a rewrite with traefik which would be instant.

Good luck

[psoliveira80]

Well, I really couldn’t solve the problem with the scenario I was faced with. What I did was replace the Traefik proxy with Caddy, but kept the N8N approach with Docker Swarm and Redis, and separated the editor, webhook, and workers. Apparently, it was a Traefik issue. I’ll still do more testing, but I think this is the way to go.