Considerations and Limitations

When you plan to deploy and configure the solution, keep in mind the following limitations and considerations.

Backup Infrastructure

When configuring the backup infrastructure, consider the following:

• You cannot connect the same Red Hat OpenShift Virtualization cluster to multiple backup servers. However, you can connect one backup server to multiple clusters — to do that, deploy the Veeam KubeVirt Proxy component in each cluster whose VMs you plan to protect.

Keep in mind that you can deploy only one Veeam KubeVirt Proxy instance per cluster.

• To establish a secure connection with a Veeam KubeVirt Proxy, you must use a trusted certificate issued by a certificate authority (CA). To learn how to configure security certificates, see OpenShift documentation.

• By design, Veeam Backup & Replication connects to Veeam KubeVirt Proxies through HTTP-based routes. It is recommended that you do not use any other connection methods.

• After you deploy the Veeam KubeVirt Proxy in a cluster namespace, do not delete this namespace and do not modify service resources in this namespace — specifically, do not change the service account that was used to deploy the Veeam KubeVirt Proxy. Otherwise, Veeam Backup & Replication will lose connection to the cluster, and you will need to add this cluster to the backup infrastructure again.
• You can protect the same cluster with both Veeam Plug-in for KubeVirt and Veeam Kasten. However, it is recommended that you do not protect the same VMs using both solutions to avoid issues with conflicting entities in the cluster.

Storage

When choosing persistent volumes that Veeam Plug-in for KubeVirt will use to store backed-up data, consider the following:

• Veeam Plug-in for KubeVirt supports persistent volumes with the Block volume mode only. The Filesystem mode is not supported.
• Veeam KubeVirt Proxy uses annotations of storage classes and volume snapshot classes to select the classes for backup and restore operations — to do that, annotate storage classes as described in section Installing Veeam KubeVirt Proxy.

VM Backup

When backing up VM data, consider the following:

• Veeam Plug-in for KubeVirt does not support backup of VMs with TPM or EFI persistence enabled.
• To ensure better performance, it is recommended that you distribute VMs evenly across Red Hat OpenShift Virtualization clusters and namespaces.
• Veeam Plug-in for KubeVirt supports only persistent volumes that use the Block volume mode — the Filesystem mode is not supported.
• Veeam Plug-in for KubeVirt does not support backup of VMs with shared volumes.
• Veeam Plug-in for KubeVirt does not support guest processing.
• Veeam Plug-in for KubeVirt does not support backup of VMs with persistent volume claims (PVCs) whose names are longer than 128 characters.

• Veeam Plug-in for KubeVirt does not support backup of VMs in UEFI mode with secure boot active.
• Veeam Plug-in for KubeVirt does not support backup of attached resources, such as service accounts, secrets and ConfigMaps.
• Veeam Plug-in for KubeVirt does not support backup of data stored on volumes attached as CD-ROM or volumes with the ReadOnlyMany access mode.

• Veeam Plug-in for KubeVirt does not support backup of VMs residing in  Red Hat OpenShift Virtualization clusters running the upstream Kubernetes.
• Veeam Plug-in for KubeVirt supports backup of VMs residing in clusters deployed using OpenShift Virtualization Operator only.
• Veeam Plug-in for KubeVirt does not support backup of VMs residing in Red Hat OpenShift Local clusters — OpenShift Local is intended for development and testing only.

VM Restore

When restoring VM data, consider the following:

• Veeam Plug-in for KubeVirt does not support restore of labels and annotations from Red Hat OpenShift Virtualization clusters to the clusters of other platforms.

• When restoring a VM from another platform to Red Hat OpenShift Virtualization, Veeam Plug-in for KubeVirt converts the disk controllers of this VM from SCSI to SATA to prevent boot failures. To ensure optimal VM performance, it is recommended that you install and enable VirtIO drivers on Windows-based VMs and VMs running minimal or customized Linux images (such as Alpine Linux), and switch the bus type to VirtIO after Veeam Plug-in for KubeVirt completes the restore operation.
• Veeam Plug-in for KubeVirt does not support Instant VM Recovery from another platforms to Red Hat OpenShift Virtualization — you can only perform Instant Recovery of Red Hat OpenShift Virtualization VMs to VMware vSphere, Microsoft Hyper-V, Nutanix AHV and Proxmox VE environments. For more information, see Performing Instant VM Recovery.

• If a restore operation requires removing an existing VM (for example, when restoring a VM to the original location or overwriting a conflicting VM), the disks attached to this VM are only removed from the cluster if they have a direct ownership link (ownerReference) to that VM. Otherwise, these disks are retained in the cluster as orphaned resources. For more information on disk ownership, see Kubernetes documentation.
• When restoring VMs with the BitLocker security feature enabled, you must enter the BitLocker keys to decrypt the volumes after the restore operation.
• [Applies to restore of Linux-based VMs] If the host name is not provided in the VM specification or the /etc/hostname file of the guest OS, Veeam Plug-in for KubeVirt sets the hostname of the restored VM to the name of the VM resource.