Your server holds a lot. It holds the warehouse and your connections. It holds installed recipes, the audit history, and stored files. All of it can be saved as one locked file. That file is called an archive. It can be loaded onto any server. You do both jobs from the webclient. Neither job needs a terminal on the server machine.
#Back up
Open Settings → Backup & Recovery. It saves the whole server. That
means the database, the settings, the secret keys, and the stored files.
All of it goes into one .recued.archive file. You download it in your browser. The
export streams as it goes. So a warehouse of many gigabytes backs up fine
on a small machine. It does not run out of memory. The server CLI can also
write the archive to a path. Use that for scripted backups.
The archive is encrypted. Your recovery key is what opens it. The recovery key is 24 secret words. Keep the archive file and the recovery key apart. Then you hold a complete restore.
The archive holds the server as it was at export time. Anything that happened after that is not in it. So back up on a rhythm that matches how much you are willing to lose.
#Restore onto a fresh server
Restoring is the first thing the webclient can do on a new server. You do not pair first and restore later. The steps:
- Install a fresh server. Open the webclient's pair form.
- Choose Restore a backup. Give it four things. The server address. The fresh server's pairing code. The backup's 24-word recovery key. And the archive file.
- The archive uploads. Then it is checked in a dry run. Nothing is saved yet. Is the recovery key wrong? It just bounces back. You can try again. The server stays untouched.
- Look at the preview of what the archive holds. Then confirm.
- The restore is saved. The server restarts into the restored data. Your webclient reconnects, and it is already paired.
A service manager, like the start-at-login setup the install script makes, restarts the server by itself. A server running in a plain terminal window needs starting again by hand.
#Restore onto a running server
A server that is already paired and in use can also load an archive. Open Settings → Backup & Recovery. Two guards apply:
- An archive from a different server identity does not quietly replace this server's data. It also needs this server's own current recovery key, plus a clear yes from you.
- An archive made by a newer server version is refused. You are told to update first. An archive from the same or an older version is fine. It is moved forward by itself at the restart.
#Built not to corrupt
- The archive is checked before anything that cannot be undone. The archive key is proven in the dry run before anything is saved.
- The restored database is put next to the live one. They are swapped in one move at restart. If a restore is cut off, the server starts on the old data, untouched.
- An upload or a save can lose its network connection. Nothing gets broken, and you do not have to upload again.
#Related
- Install the server — where the data directory lives.
- Pairing and devices — the recovery key this flow depends on.