Skip to content
Cannot access n8n via SSH or web console — backup needed before recovery
Anonymous
OP

a month ago

I cannot access my Railway n8n service through either the web console or SSH.

Web console:

  • “WebSocket connection failed”
  • The file browser also shows “Connection failed / Connection lost”.

Railway CLI:

  • “kex_exchange_identification: read: Connection reset”
  • “Connection reset by 66.33.22.3 port 22”

SSH previously worked with the same registered key. Subsequent attempts fail before authentication. I have also tried another Wi-Fi network.

Context:

A rebuild upgraded n8n from 2.35.7 to 2.42.3 because the Dockerfile uses n8nio/n8n:latest and USER root. n8n generated a new encryption key in /root/.n8n/config and cannot decrypt the existing credentials.

A previous inspection confirmed that the persistent volume is mounted at /home/node/.n8n, but that directory has no config file.

Workflows are paused. I need to preserve the current container configuration and back up Postgres before continuing recovery.

What diagnostics or supported access alternatives can I use without restarting, redeploying, or replacing the current container?

I am not assuming this is a platform-wide outage.

Captura de pantalla 2026-10-06 202019.png

$10 Bounty

1 Reply

Status changed to Awaiting Railway Response Railway • about 1 month ago


Railway
BOT

a month ago

This thread has been opened as a bounty so the community can help solve it.

Status changed to Open Railway • about 1 month ago


hibaech1t

a month ago

Since the priority is preserving the current state without restarting or redeploying the n8n container, have you tried using Railway’s volume file tooling to inspect/download the persistent n8n volume directly? Railway currently documents railway volume browse / and railway volume files download for accessing files stored on a volume. This could allow the /home/node/.n8n volume contents to be preserved without modifying the running n8n deployment.

For PostgreSQL, Railway also supports creating a manual volume backup from the Postgres service’s Backups tab. That would preserve the database state before attempting recovery.

Since the missing encryption key may exist only in /root/.n8n/config, I would avoid restarting or redeploying the current n8n container until Railway confirms whether that non-volume path can be recovered through another supported access method.


Welcome!

Sign in to your Railway account to join the conversation.

Loading...