Nothing here yet.
Nothing here yet.
No publications yet.
That engines-field detail is the part I would check first. Running node -v before installing, and pinning the Node version, stops npm quietly settling for an older n8n. For the always-on step, I would also test one webhook end to end after moving it, since a local URL that worked on your PC is the usual thing that breaks on a server.
The silent fallback to an old n8n version is the trap worth warning people about — no error, just an old install. Backing up .n8n\config together with database.sqlite is the detail that saves your credentials later. If you ever need webhooks reachable from the internet, a one-click n8n deploy on RushDeploy keeps it online without keeping your PC on — the free plan covers 1 project and 50 deploys a month.
The restore test is the part most people skip, and checking that credentials still decrypt after the restore is the detail that makes it real. One thing I'd add is a longer-lived monthly snapshot alongside the 30-day rotation, since a bad workflow edit can sit unnoticed for weeks. That way a restore point from before the problem still exists.
The step most people skip is the post-move sweep of hardcoded URLs in widgets and custom fields. A serialized-safe search-replace fixes the database, but a quick crawl of the new site for any remaining old-domain strings catches what the plugins leave behind. Running that crawl before removing the noindex is the cheapest insurance in the whole sequence.