Backup & restore
Your whole site is one SQL file (.pod, a plain SQLite database, only renamed). Back it up or move it to another server from the admin in two clicks.
Download a backup
Settings → Pod → Backup & restore → Download SQL file (admins only). You get a consistent copy of the live database as content-<date>.pod, without active sessions. Media stored in the pod are inside the file; media on S3 are not.
Automatic backups
The admin copies the pod into a backups/ folder next to it, by default every 7 days (checked at startup and every six hours), and keeps the newest four. Back up now in the same card does it on demand. Tune it with environment variables:
| Variable | Default | Meaning |
|---|---|---|
ORBITER_BACKUP_DAYS | 7 | Days between automatic backups, 0 turns them off |
ORBITER_BACKUP_KEEP | 4 | How many automatic (and manual) copies to keep |
ORBITER_BACKUP_DIR | <pod folder>/backups | Where the copies go |
ORBITER_POD_WARN_MB | 60 | Size above which the dashboard and the card warn |
Keep the backup folder private. A backup holds password hashes and saved settings. Since admin 0.3.105 the files are written with mode 600 in a 700 folder. Still prefer a folder outside any directory your web server serves (ORBITER_BACKUP_DIR), and check that https://your-host/backups/ does not answer. Older versions wrote the files world-readable.
With automatic backups off, the dashboard reminds you when the last backup is older than two weeks. A copy on the same disk protects against mistakes, not against losing the server: download one now and then.
Keeping the pod small
The card shows the file size and how much of it is media stored inside the pod. If you want to stay below a limit, choose a media backend under Settings → Media (local folder, S3 or GitHub) and press Move media out of the pod…: the files are uploaded one by one (a file that fails stays where it is), then the pod is shrunk with VACUUM. Make a backup first. Moving to S3 or GitHub needs working credentials; the local folder is the simplest.
Restore or move
Restore… takes a .pod file. The server checks it first (SQLite, integrity check, required tables, at least one admin) and refuses anything else. It then keeps the current pod next to it as <pod>.before-restore-<time> (the last three are kept), swaps the new file in and signs you out.
- Credentials saved encrypted need the same
ORBITER_SECRETon the new server. - A pod from Lander works too; its admin passwords are PHP hashes, so reset the password with the CLI if you cannot sign in.
- The upload size is capped by
ORBITER_MAX_UPLOAD_MB(default 100).
.pod to a public repository.API
GET /api/backup — download the pod (admin) POST /api/backup/restore — multipart field "file" (admin) GET /api/backup/status — size, last backup, automatic backups POST /api/backup/run — back up now into the backup folder POST /api/backup/offload-media?limit=25 — move media out of the pod POST /api/backup/vacuum — shrink the pod file