This episode traces the origins of Proxmox Backup Server: a project started by Proxmox in 2018 with the first stable 1.0 release in 2020 as the backup backbone of the Proxmox VE ecosystem. You will also understand why PBS is needed — per-chunk deduplication, client-side encryption, instant restore, native integration, off-site sync, and its role as a Veeam/Bacula target.

With the environment ready in episode 0, now we talk about why PBS exists and why you need it. Understanding the background does not only add insight — it explains why PBS is designed the way it is: per-chunk dedup, client-side encryption, and seamless integration with Proxmox VE.
Imagine PBS like a warehouse with a smart indexing system: it does not re-store every identical document, it stores each once and records who uses copies of it. It also locks every envelope with a key only the sender holds, so even the warehouse staff cannot read it. That is the essence of the PBS philosophy: efficient, secure, and fast.
PBS is developed by Proxmox — the same company behind Proxmox VE and Proxmox Mail Gateway. The project started in 2018 as an answer to a backup need that existing solutions did not fulfill, and the first stable release 1.0 shipped in 2020. PBS was born as the backup backbone of the Proxmox VE ecosystem: it understands the data structure of VMs and containers deeply, rather than merely copying raw files.
Its philosophy is simple yet firm:
PBS's journey has been quite active: PBS 3.x (2023-2025) introduced namespaces, 3.4 (April 10, 2025) closed the 3 series, 4 (August 6, 2025) brought a new Debian base, and 4.2 (April 29, 2026) is the current release with S3 object storage support. The release model follows Debian, and PBS 3 security support ends August 31, 2026 — discussion material for episode 17.
The biggest problem with VM backups is storage waste: hundreds of VMs run the same operating systems, storing the same data over and over in every snapshot. PBS splits data into chunks and performs content-defined dedup, so a single block of data is stored only once even if it appears in dozens of backups. Compression (zstd by default) squeezes the size further. The result: disk requirements can drop drastically, and backup bandwidth shrinks.
Backup data is an easy target for thieves. PBS enables client-side encryption: data is encrypted before it is sent to the server, so the server — even the PBS admin — only stores ciphertext. Without a key stored safely on the client, nobody can read the snapshots. Practical details in episode 7.
PBS restore is very fast because it uses the sparse mechanism: blocks containing zeros are not allocated, and restore can begin while data is still streaming over the network. You can boot a VM from the backup as if from the original disk, without waiting for the full restore to finish — crucial during incidents and when RTO is tight.
This is PBS's biggest differentiator. In a Proxmox VE datacenter, PBS is registered as a one-click backup storage: the PVE scheduler calls PBS natively, restore can be done directly from the interface, and file restore is available per VM. No glue scripts or manual import/export.
PBS provides a sync job that pulls snapshots from one PBS server to another PBS server in a different location. The combination of dedup + remote sync makes the 3-2-1 strategy (3 copies, 2 different media, 1 off-site) affordable — we practice this in episode 10.
PBS is not selfish: its datastore can be exposed as an NFS/SMB share, so other enterprise backup tools such as Veeam or Bacula can use PBS as a repository target. This expands PBS's role from "Proxmox backup" to "backup storage engine". Details in episode 19.
Key takeaways:
In the next episode, episode 2, we will break down the core concepts & key PBS architecture — what a datastore is, how content-defined chunking works, what backup groups and snapshots are, and the role of components such as proxmox-backup, the web UI on port 8007, proxmox-backup-client, and the API. This foundation will make all the following hands-on episodes feel light!