Environment
- n8n version: 1.93 (self-hosted, managed hosting)
- n8n-nodes-zalo-tools version: 0.6.1
Bug description
Every node from this package fails immediately with a generic "Zalo connection failed" error, no matter which node or which workflow. No execution is saved to the Executions history for these runs.
Steps to reproduce
- Create a brand-new, completely empty "Zalo API" credential (Cookie/IMEI/User Agent/Proxy fields all blank — never used before).
- Add a Zalo Login Via QR Code node, assign the new credential (plus a valid n8n Account credential with apiKey+url), click "Execute step".
- Result: fails in under 3 seconds with "Zalo connection failed" — no QR code is ever generated.
- Same result on Zalo Message Trigger — "Test this trigger" fails identically in under 3 seconds, with no output.
- Same result on Zalo Group (getAllGroups) and Zalo Send Message.
What we've ruled out
- Not a stale/expired session — reproduced on a completely fresh, never-used credential.
- Not an n8n API key issue — regenerated it, no change.
- Not general outbound network failure — an HTTP Request node on the same instance successfully calls https://httpbin.org/get.
- Not a workflow-design issue — reproduced identically across two independently-built workflows.
- Not a credential-reference issue — verified via n8n API that each node correctly points to the new credential's ID.
Suspected cause
Matches n8n core issue n8n-io/n8n#15620 (n8n-io/n8n#15620) ("Zalo trigger fail", closed as not planned, no resolution). Guessing this package calls back into n8n's own internal REST API (e.g. to persist a refreshed Zalo session cookie) and that callback is refused.
Question
Is there a known networking requirement (e.g. a specific internal host/port) this package expects to reach from within the n8n process itself? Any detail on what "Zalo connection failed" means internally would help.
Environment
Bug description
Every node from this package fails immediately with a generic "Zalo connection failed" error, no matter which node or which workflow. No execution is saved to the Executions history for these runs.
Steps to reproduce
What we've ruled out
Suspected cause
Matches n8n core issue n8n-io/n8n#15620 (n8n-io/n8n#15620) ("Zalo trigger fail", closed as not planned, no resolution). Guessing this package calls back into n8n's own internal REST API (e.g. to persist a refreshed Zalo session cookie) and that callback is refused.
Question
Is there a known networking requirement (e.g. a specific internal host/port) this package expects to reach from within the n8n process itself? Any detail on what "Zalo connection failed" means internally would help.