Unexpected Workflow Data Rollback on Multiple Independent n8n Instances (DigitalOcean)

Questions
1

I’m facing a serious issue with data loss on two completely separate n8n instances hosted on DigitalOcean droplets, and I’m trying to understand what might be causing it.

I have two droplets:

  1. Personal n8n instance – one VPS

  2. Work-related n8n instance – a different VPS, different project, no shared resources

These droplets are not connected in any way — different accounts, different setups, different workflows.

However, the same problem happened on both:

  • On the first droplet, all changes made during the last month were rolled back. All workflows created or modified during that period simply disappeared.

  • A few days later, on the second droplet, the system rolled back about 24 hours of changes — again removing new or updated workflows.

There were no manual restores, no backups applied, and no migrations. The containers were not recreated with docker compose down -v. The Docker volumes were intact. Logs do not show any obvious errors.

This looks like an unexpected revert to an older database state, but I have no idea what could trigger it.

Has anyone encountered similar behavior?
What could cause n8n (or its database) to revert to an old state like this across completely independent servers?

Any insights or hypotheses would be appreciated.

2

Hey @rustamguloff Welcome to the community.

Really unfortunate and I’ve never experienced anything like that in my 2 years of working with n8n. Did the version of n8n change as well or just the data? It’s been a while since I used DO so I’m not sure if this was something like a DB snapshot revert…

Don’t really have an answer but just wanted to take the opportunity to recommend/remind people that you need a templates backup workflow running in the background Discover 7164 Automation Workflows from the n8n's Community. In situations like these, it can mean a much faster and less expensive recovery.