Static demo — actions are disabled. Source and install
TasmoFleet
Snapshots , 1 new since last visit
TasmoFleetv0.36.0
Account · maren What’s new 1 new
Theme

Snapshots

Copies of TasmoFleet’s own database. Device backups are a separate thing.

Download a snapshot

Acts at once, outside the schedule form. Every download is written to the audit log.

A consistent copy of the live database, made while TasmoFleet runs. You are asked to confirm that the file holds secrets.
The file holds secrets. Keep it like a password.
  • Inside: devices, groups, settings, users with their password hashes, sessions, audit log and the backup index.
  • Encrypted inside: device HTTP and MQTT passwords, the default device password, the MQTT broker password, the OIDC client secret and refresh tokens, the metrics token, the notification webhook header value and SMTP password. The key is secret.key and it is not in the database file: without it these values cannot be read, with it they can.
  • Not inside: device backup files (.dmp), the firmware cache, secret.key (unless ticked above).

Stored snapshots

Newest first, in /data/snapshots. Delete acts at once. File names are in UTC (the Z): tasmofleet-20261106-0800Z.db was taken at 09:00 in Central European Time (10:00 in summer time). The times in this list follow your browser's clock.

No snapshots yet. Choose days on the schedule on the right, or download one now. Downloads are not kept on the server.

Schedule

Every scheduled snapshot is written to the audit log. The failure message is set under Settings → Notifications.

Days
None ticked = off.
Time of day on the server's clock (TZ). A snapshot missed while TasmoFleet was down is taken late within 60 minutes; after that it is skipped and a notification is sent. Saving the schedule never takes one.
On the server's clock (TZ).
The oldest are deleted first.

Times are on the server's clock (UTC).

Stored in the data directory (/data/snapshots), so the host backup of /data covers them. They sit on the same disk as the database: they protect against a bad upgrade or a deleted row, not against a dead disk.

Restore

By hand, with the container stopped.

  1. Stop the container.
  2. Replace /data/tasmofleet.db with the snapshot and delete tasmofleet.db-wal and tasmofleet.db-shm.
  3. Make sure /data/secret.key is the one from the same time, then start the container.

There is no restore button on purpose. A snapshot from an older version is migrated at start; one from a newer version is refused. The tasmofleet.db.pre-N copies made before an upgrade are restored the same way.

No unsaved changes Discard