XCP-ng Business Continuity: Backup and Disaster Recovery

One approach that works is image based backup at the hypervisor level. It protects whole virtual machines with no agent inside the guest. With a forever incremental strategy, only the first backup is full and every later run picks up the blocks that changed. Changed Block Tracking (CBT) makes those incremental jobs faster and easier on storage. Deduplication, compression and skipping blocks that do not need to be stored cut the storage use further.

There are several ways to lay out the infrastructure. Where there is SAN storage, LAN-free backup sends the data over the storage network instead of the production LAN, so users notice less. Without a SAN, a dedicated backup network keeps that traffic separate. Distributed backup nodes let the setup grow as the VM count and the production data grow.

Disaster recovery planning matters just as much. An offsite copy replicates the backups to a second site so you can recover quickly when the primary site goes down. A cloud archive covers long-term retention and the case where something physical happens to the site.

On the recovery side, granular file-level restore saves you from recovering a whole VM when someone needs one file. Instant restore powers critical workloads on straight from the backup storage, which cuts downtime sharply. Cross-platform recovery helps when you are migrating as well.

Efficient backups, a deployment model that scales, and several layers of recovery together are what keep an XCP-ng environment available.