FAQ — Backups
Which resources can be backed up?
PostgreSQL and MariaDB databases, virtual machines (with all their disks) and disks. MongoDB, Redis, RabbitMQ, Kubernetes clusters and buckets are not covered by this service. The data of a Kubernetes cluster is backed up with tools installed in that cluster.
Where are my backups stored?
In the S3 bucket of your project chosen for the destination. Hikube keeps no other copy: deleting the bucket deletes the archives.
Do backups have a quota?
No quota of their own. Archives use the space of their bucket, counted like the rest of object storage.
Do I need to create one destination per resource?
No. A destination serves all the covered resources in the project. Create several for different retentions or buckets.
Can I back up a PostgreSQL database to two destinations?
No. A PostgreSQL database continuously archives its transaction logs to a single destination; two destinations would share this stream and neither would contain a complete history. The second one is refused. Several schedules to the same destination remain possible. See Concepts.
Which time zone are schedule times in?
UTC. The console preset converts your local time to UTC; the cron expression is entered directly in UTC. A scheduled backup therefore shifts by one hour in local time at each daylight saving change.
Can I back up more often than once an hour?
Not with a schedule: the maximum frequency is one run per hour. For PostgreSQL, it is not necessary: transaction logs are archived continuously, and point-in-time restore reaches any point between two backups.
How do I get a copy of a database without touching the original?
Create an empty database of the same type (and of the same major version for PostgreSQL), then restore the archive onto it with Another application. See Restore a database.
Can I restore a VM under another name?
No. A VM or a disk is only restored onto itself. To recover files without replacing the whole VM, restore it choosing Keep them: the old disks remain available as separate disks. See Restore a VM or a disk.
Can I restore a single disk of a VM?
No. A VM's archive is restored as a whole, with all its disks. The archive of a disk backed up on its own is only used to recreate it if it has been deleted.
Does deleting an archive free up space in my bucket?
For a VM or a disk, yes: the data is erased shortly after deletion. For PostgreSQL and MariaDB, no: only the record disappears from the console, the data stays in the bucket.
What happens to my archives if I delete the backed-up resource?
They are kept and remain restorable. For a VM or a disk, the restore recreates the resource; for a database, choose another database as the target.
What happens to my archives if I delete the destination?
They stay in the bucket, but can no longer be restored from the console, even if you create a destination of the same name again. For PostgreSQL, their retention is no longer enforced.
Why is the size of my archives not shown?
It is not measured today. For a VM or a disk, the percentage shown during a backup measures the data transferred, not the space taken up. The actual space can be read on the bucket.
What happens if a scheduled backup falls during a restore?
An on-demand backup is refused during a restore, but a scheduled backup still triggers. For a PostgreSQL database restored in place, it can make the restore fail: before restoring, check that no run is planned within the next 30 minutes.