Waiting webhook links use localhost in “Send and wait” emails despite correct WEBHOOK_URL and N8N_EDITOR_BASE_URL

Questions
1

Describe the problem/error/question

In production, the “Send and wait” (Gmail node) generates waiting webhook links that point to http://localhost:5678/webhook-waiting/... instead of my public URL https://n8n.pulseflow.ch.

All other webhooks display correctly (trigger functions etc.)

What is the error message (if any)?

No error message, but the approval/response link in the email is invalid because it points to localhost.

Example link from email:
http://localhost:5678/webhook-waiting//?approved=true&signature=…

Expected:

Please share your workflow

Here is the Gmail node config (simplified):

{
  "nodes": [
    {
      "parameters": {
        "operation": "sendAndWait",
        "sendTo": "[email protected]",
        "subject": "Approval",
        "message": "Approval email content...",
        "responseType": "customForm",
        "formFields": {
          "values": [
            {
              "fieldLabel": "Confirm?",
              "fieldType": "dropdown",
              "fieldOptions": {
                "values": [
                  { "option": "Yes" },
                  { "option": "No" }
                ]
              }
            }
          ]
        }
      },
      "id": "a00ec081-07c0-4f8a-9139-6606a65236ff",
      "name": "Send Content Confirmation",
      "type": "n8n-nodes-base.gmail",
      "typeVersion": 2.1
    }
  ]
}

Share the output returned by the last node
Email is successfully sent, but the approval link is broken (points to localhost).

Information on your n8n setup
n8n version: 1.114.4

Database: Postgres

n8n EXECUTIONS_PROCESS: queue (workers enabled)

Running via: Docker behind reverse proxy (TLS terminates at proxy)

Operating system: Linux (containerized)

Relevant environment variables (set on ALL containers: main + workers):

N8N_EDITOR_BASE_URL=https://n8n.pulseflow.ch
WEBHOOK_URL=https://n8n.pulseflow.ch/
N8N_HOST=n8n.pulseflow.ch
N8N_PROTOCOL=https
N8N_PORT=5678   # proxy handles 443
GENERIC_TIMEZONE=Europe/Berlin
N8N_BLOCK_ENV_ACCESS_IN_NODE=true
N8N_ENCRYPTION_KEY=...

Diagnostics
/rest/settings confirms correct values:

urlBaseEditor: https://n8n.pulseflow.ch

urlBaseWebhook: https://n8n.pulseflow.ch/

endpointWebhookWaiting: webhook-waiting

Proxy forwards Host and X-Forwarded-Proto.

Restarted all containers and re-triggered new executions.

Verified envs are present in both main + worker containers via printenv.

Issue persists: email links still show http://localhost:5678.

Question
In queue mode, do workers need an additional env var (besides WEBHOOK_URL / N8N_EDITOR_BASE_URL) for waiting link generation?

Is this a bug in how workers build sendAndWait webhook links?

Any known fixes/workarounds for this?

Thanks a lot 🙏
2

Hi @RainyDev

According to the docs:

  • Set the N8N_PROXY_HOPS environment variable to 1.
3

Hi @mohamed3nan

Thanks for your quick response!
I set the ENV, unfortunately still the same issue.

Somehow its only the /webhook-waiting Webhook im facing this issue with. All other sorts of webhooks get displayed correctly which adds some confusion to me.

Any other ideas?

Best regards

4

Please have you been able to fix this error?

5

I also have this when using the “Send and Wait for Response” operation on the “Send a text message” Node of Telegram.

Very confusing as my WEBHOOK_URL is properly set and works for webhook nodes.

N8N_PROXY_HOPS’s value is 1

6

Could anyone fix this in the meantime?

Also facing the exact same issue - it works with regular webhook nodes but e.g. not with Gmail, Slack etc (hint: Redirect URLs are added there)