| Version compatibility for clusters and virtual machines |
The VMware vSphere version of the source cluster, the firmware type and operating system version of the virtual machines to be migrated, and the Arcfra Cloud Operating System (ACOS) version of the destination cluster must be supported in Compatibility.
|
| Virtual machines to be migrated |
- The virtual machine is in the
Running or Shutdown state.
- The operating system is installed.
- The number of snapshots for a single virtual machine is less than 32.
- The PCI passthrough or PCIe passthrough is not configured.
- The number of virtual disks on a single virtual machine is not 0.
- The virtual disk file is in
.vmdk format.
- For the virtual machine managed by backup software featuring virtual machine snapshots, you need to stop the corresponding backup task first before starting the migration.
- For the virtual machine with the RDM disk, the RDM disk is unmounted, or the virtual machine is shut down before migration.
- For the virtual machine with the disk whose
Sharing attribute is Multi-writer, the Sharing attribute is changed to the No sharing before migration.
- For the virtual machine with the disk whose
Disk mode is Independent - Persistent or Independent - Nonpersistent, the Disk mode is changed to Dependent, or the virtual machine is shut down before migration.
- For the virtual machine using the Linux operating system and with SCSI and IDE disks, and the disks are listed in the
/etc/fstab file of the virtual machine's operating system using drive letters, the drive letters are changed to the disk UUID before migration.
|
| Destination cluster resource space |
- The available number of CPUs in the cluster is sufficient to support the normal operation of destination virtual machines.
- The available memory space in the cluster is greater than the total memory space of all virtual machines to be migrated.
The available data space in the cluster is greater than the estimated data space needed.
The data space calculation formulas for different storage policies are as follows:
- The redundancy policy is replication:
- For the virtual machine adopting thin provisioning as the storage policy: Estimated data space = (Sum of all virtual disk sizes of virtual machines to be migrated + 5GiB) × Number of replicas
- For the virtual machine adopting thick provisioning as the storage policy: Estimated data space = (Sum of all virtual disk sizes of virtual machines to be migrated + Sum of all system disk sizes of virtual machines to be migrated) × Number of replicas
- The redundancy policy is erasure coding (k + m):
- For the virtual machine adopting thin provisioning: Estimated data space = (Sum of all virtual disk sizes of virtual machines to be migrated + 5GiB) × (k+m)/k
- Provisioning method is thick provisioning: estimated required data space = (sum of virtual disk sizes of all virtual machines to be migrated + sum of system disk sizes of all virtual machines to be migrated) × (k+m)/k
Note that the virtual disks mentioned above include data disks and system disks. At this stage, since it is unknown which disk is the system disk, it is recommended to take the average size of all virtual disks in a virtual machine as the system disk size.
|
