VirtConsole 3.3.3 lets customers restore a managed database backup into a brand-new database, helps administrators move reverse DNS over from another panel, and makes backups and storage readings more dependable. 3.3.3 is available now under updates in your panel.
Restore a database backup into a new database
A full VM backup of a managed database can now be restored into a new database in the same location, from the panel or the API. The original database is left untouched, and the copy keeps the same engine and version.
- The new database is billed from the moment it becomes active, never while it restores. If a restore fails, nothing is charged.
- The copy never touches the original’s backups. Before its first boot, the node removes the source’s backup credentials from the disk, and it refuses to boot if that clean-up fails.
- Until the new database is activated, it cannot be redeployed, suspended, resumed or reinstalled, and the backup it is restored from cannot be deleted.
- A backup whose database major version has since changed can now be restored this way, instead of being refused.
- Database detail pages show where a restored database came from.
Reverse DNS: adopt a zone and import its records
If you are moving from another panel, your reverse zones probably already exist on PowerDNS or ClouDNS. You can now adopt an existing zone without rewriting its NS or SOA records, preview exactly what an import would do record by record, and then import the PTR records as a background task. The import only reads from your DNS provider and never writes to it, and by default it fills only addresses whose reverse DNS is still empty. Deleting an adopted zone in the panel never deletes it at the provider.
Recover a backup stuck in deletion
A backup on a local, S3 or rclone destination could stay marked “being deleted” if the node never confirmed the delete. Administrators can now resolve it through a guided recovery: request it from the backup page, drain the node and run one command on it. The node reads the backup file itself. If the file is there, the backup is kept. If it is provably gone, the record is removed. Nothing is deleted automatically and nothing is released just because time has passed.
Backups that cannot overlap
Full and incremental backups of one managed database now go through a single check, so two never run at once, and a replica taking a policy backup holds its whole family. A backup stuck in progress can be closed by an administrator, with a clear message that nothing is deleted from the backup destination.
Storage health in more detail
- Read-only, out-of-space, failed or check-needed LVM thin pools are flagged.
- ZFS shows the space datasets can actually use.
- Every reading shows when it was last attempted and is marked stale when it is out of date.
Also in this release
- Reverse DNS records are written to PowerDNS as absolute names.
- Kubernetes regions and VPC pickers list only locations that can actually host a VPC.
- Console sessions and several backup, restore and delete paths are more reliable, with clearer status codes and reasons.
- Deleting a user is refused with a clear reason while they still own VPCs, NAT gateways, VPN gateways or static IPs.
Thanks for building on VirtConsole. Questions along the way, reach us any time.