Harvester VM restore: disks not reattached automatically

Problem Overview

When restoring a VM that had multiple data disks attached, the disks are recreated but are not automatically reattached to the VM. The network interface also reverts to default settings, potentially changing the IP address.

Goal and Constraints

We need to understand the exact restore behavior for multi‑disk VMs and verify that the restored disks contain the same data as the originals. The restore operation uses Velero snapshots and the harvester vm restore command, which offers a --verify flag for checksum validation. However, it is unclear whether this flag covers all disks and whether network settings are preserved.

Unresolved Questions

1. Does harvester vm restore automatically reattach all disks that existed before the snapshot, or must they be reattached manually?
2. When a VM is restored, do the network interfaces retain their original IPs, or are they reset to defaults?
3. Is the --verify flag sufficient for ensuring application‑level data integrity, or should additional checks be performed?