Ceph under Hyper-V and VMware
Our ingenious embedded solutions give you an economical and highly secure backup repository.
Ceph supports object and block storage as well as the POSIX file system in a cluster. Depending on their backup requirements, customers can select different storage protocols to support the requirements of different storage backup strategies. Here we have chosen a Ceph block storage (Ceph RBD) and a Ceph file system (CephFS) as a backup partition.
Architecture of a backup solution with Veeam and Ceph
The architecture of virtual machine backups under VMware and Hyper-V are similar. Veeam uses the so-called data mover for this. So it transfers the data from the source hosts to the backup repositories. Incidentally, the data mover runs on both the proxy server and the repository server. To use Ceph as backend storage for a backup partition, the administrator can mount RBD or CephFS on a physical Linux server or a VM as a backup server.
If the proxy and repository server are virtual machines within the hypervisor cluster, you can take advantage of network-free high-speed data transport between the VM disk, proxy server and storage target. Tests have shown that the best configuration of a large hypervisor cluster is to provide a proxy server VM and a repository server VM on each VMware host.
Otherwise, you can deploy a backup proxy VM on each VMware host as well as an off-host partition VM. This also relieves the workload of your productive VMware host.
Flexible VM backup with Veeam and Ceph
There are three options, one Ambedded Ceph appliance as a repository for Veeam backup and replication to use:
- RBD block devices
- S3 object storage
Whereby CephFS – and RBD block devices can also be used as local backup media. The S3 object storage serves as a capacity tier for a remote location. For more information, see the Ambedded White Paper.
He is our system engineer for Linux, Ceph and ZFS.