Step 1 The client on the Hyper-V node where the target VMs reside invokes the WMI interface to create a VM snapshot.
Step 2 The client obtains the disk path of the VM snapshot (checkpoint), and mounts the virtual disks corresponding to the VMs to the Hyper-V node.
Step 3 Parse the changed data block information of the differencing disk (avhd/avhdx file) between the latest snapshot (checkpoint) and the snapshot (checkpoint) retained by the last backup, used to calculate the data blocks to be backed up for incremental backup. For full backup, this step is skipped.
Step 4 The client reads the data in the virtual disks mounted to the Hyper-V node and backs up the VM configuration file vm.xml or vm.bin, m.vsv.
Step 5 After the backup is completed, delete the last snapshot (checkpoint) and keep a copy of the latest snapshot (checkpoint).