Hello,

I'm experiencing a critical issue with my n8n service and need help recovering access to existing data before it's potentially lost.

PROJECT DETAILS:

• Service: n8n (self-hosted)
• Domain: n8n-terapeuta.up.railway.app
• Project ID: 1b83b3c4-d6c6-4eff-a34f-891ecc8698eb
• Volume ID: df2691de-5fdc-4b27-ad50-ec2cf5e00f61
• Environment ID: 4db08992-691a-4bd0-8214-073d96eb507c

WHAT HAPPENED:

1. My n8n instance was working normally with multiple production

   workflows, credentials, and execution history.

2. I added an environment variable (NODE_FUNCTION_ALLOW_BUILTIN=crypto)

   and triggered a redeploy to allow the Node.js crypto module in

   Code nodes.

3. After the redeploy, the service started showing the "Set up

   owner account" screen — as if it were a brand new n8n installation

   with no existing data.

4. I checked the connected volume settings:

   • Mount path: /home/node/.n8n (correct)

   • Volume usage: 413 MB / 500 MB (data clearly still exists

     on the volume)

   • The volume is connected to the correct service

5. I removed the environment variable and redeployed again, but

   the issue persisted — still showing the fresh setup screen.

6. I checked deployment logs and saw hundreds of database migrations

   running from scratch (e.g., "Starting migration

   AddRestoreFieldsToWorkflowBuilderSession..." through very recent

   ones), which suggests the service may be reading/creating a NEW

   SQLite database file rather than connecting to the existing one

   inside the volume.

7. There is only one deployment listed (no rollback history available),

   so I could not test an older deployment version.

8. Unfortunately, I proceeded to complete the "Set up owner account"

   flow before realizing this likely indicated a fresh database, and

   am now in the n8n onboarding/customization screens.

QUESTION / REQUEST:

Could you please check the actual filesystem contents of the volume

(mount path /home/node/.n8n) to confirm whether the original

database.sqlite file (or equivalent) still exists there with my

previous workflows, credentials, and execution history?

If it does, I would appreciate guidance on how to reconnect the

service to that existing database file instead of the new one that

appears to have been created.

I have not made any further changes to the volume (no wipe, no

delete) since this started, so I believe the original data should

still be physically present.

This is time-sensitive as the n8n instance is running production

automations. Thank you very much for your help.

Best regards,

Ricardo Oguido