Platform-managed daily backups

Daily Hetzner snapshots at 03:00 server local. 7-day rolling retention. One-click restore.

Every VICIfast server gets a daily Hetzner snapshot at 03:00 server local. Seven snapshots live at any time; day 8 evicts day 1. Restore is one click from the dashboard.

What's in the snapshot

A full disk image — VICIdial DB, recordings, Asterisk config, your crontab, /var/log, everything. Snapshots are full images, not incrementals; no chain to corrupt.

Restoring

Dashboard → your server → Snapshots.
Pick a snapshot, click Restore.
Server reboots from the snapshot image. Usually 90–120 seconds of agent downtime.

What gets lost on restore

Anything written between the snapshot time and now. If the snapshot was at 03:00 server local and you restore at 14:00, the 11 hours of dialing in between is gone. This is true of any snapshot system; for point-in-time recovery use MySQL binlog shipping to your own server.

Restore to a different server

Admin-only operation. Useful for spinning up a staging copy from a prod snapshot, or migrating between regions.

Adjusting the schedule

03:00 server local is the default — quietest hour for outbound dialing in most US markets. Floors running offset hours can request a different snapshot window via support.

Restore drill

We run a weekly restore drill against sample customer snapshots in a sandbox. Failures page the on-call. The drill verifies the snapshot is actually restorable, not just present.

For longer retention

Daily snapshots roll on day 8. For 30-day / 90-day / multi-year retention, see the External Backups feature — push to your own S3 with whatever lifecycle policy fits your compliance.