A download link lets another server, or a person, download an archive without signing in. It is the simplest way to move a site: the archive goes straight from the old server to the new one, however large it is.
Making a link
Click Get download link next to a complete archive. The link appears at the top of the Backups page, once, with a Copy button. It looks like:
https://old.example.org/backup-download/12/3f9a…(64 characters)
An archive kept in cloud storage is sent through the site, so its address in the storage stays private and revoking the link stops it.
The link works for 48 hours. The list shows until when; Revoke link stops it at once. Making a new link for the same archive replaces the old one.
Only a fingerprint of the link is stored, so the link cannot be recovered from the database: if you lose it, make a new one.
Using a link
On the other site's Backups page, paste the link under Bring in an archive > From a link and click Download. The download runs in the background, continues where it stopped if the connection drops (up to five tries), and the archive is then checked file by file before it can be restored.
The same box takes any direct link to a backup archive, such as a file shared from cloud storage.
Keep links private
Anyone with the link can download the archive, which holds the whole site, password hashes and API keys included, until it expires or you revoke it. Send it only to yourself or the person moving the site, revoke it when the move is done, and use https.
Every use of a link is written to the log in the backup directory (backup-<random characters>.log) with the address it came from.
Links to private networks
For safety, the From a link box only downloads from public internet addresses. See Configuration to allow private network addresses for moves inside a private network.