Hello n8n Team/Community,
I need help analyzing and recovering data after a serious server failure (VPS Mikrus) and the loss of current data. I’ve reached a stage where the n8n container is running, but it’s not loading the saved workflows.
Configuration Details
-
n8n Version: 1.113.3 (latest available at the time of the last run)
-
Hosting Method: Docker Compose
-
Database: SQLite (database.sqlite)
-
Server: Linux VPS (Mikrus)
Situation Description and Steps Taken
-
Failure: My VPS crashed. I lost the current data (from the last week), but I had two older backups.
-
Restoration: I attempted to restore data from the oldest but only available file: n8n_backup_2025-10-05.tar.gz.
-
File Verification: After unzipping, the database.sqlite file is present in the
/root/n8n/data/directory and has a size of ~536 KB. Permissions were fixed (chown 1000:1000).
The Problem (Diagnosis)
Despite the correct upload of the database.sqlite file and the container starting successfully, n8n:
-
Does not load administrator data: When I visit the URL, instead of the login screen, I see “Set up owner account”.
-
No Workflows: After creating a new administrator account, the dashboard is completely empty (I see “everything new”). It appears n8n considers the database empty or incompatible.
Questions for the Community/Support
-
Could the issue be due to corruption in the user/authentication tables within the
database.sqlitefile? -
Is there a tool to check the integrity of my
database.sqlitefile or to extract only the workflow data from it (without credentials) for manual import? -
Could the lack of a manually set
N8N_ENCRYPTION_KEYvariable in the.envfile be causing this issue (the database used a default/old key and is unreadable by the new installation)?
If possible, I would be willing to provide the database.sqlite file for analysis to verify its content.
Thank you for any help and guidance!