Immutability, single-use and certificate-based credentials, and what a hardened repository will not let you do.
From Ultra Transcenders VBR-13 by Tony Rough (coming January 2027)
A hardened repository is a Linux server that makes backup files immutable for a set period and keeps no usable credentials on the backup server. It is the main on-premises defence against an attacker who has taken over the backup server.
It offers immutability: for the configured period, files cannot be moved, modified or deleted, although they can be copied. It also offers secure authentication. Certificate-based authentication needs no user name or password and is available on Veeam Infrastructure Appliance deployments. Single-use credentials are used once to deploy the Data Mover and never stored, so a compromised backup server cannot reuse them. No other role may run on it except a VMware backup proxy in Network (NBD) mode, and even that adds VDDK and its vulnerabilities, so Veeam recommends placing the proxy elsewhere. It can store VM, physical, cloud-copy, unstructured data, log and plug-in backups, plus VeeamZIP, Copy, Move and Export Backup output.
| Deployment option | Authentication | Points to know |
|---|---|---|
| Veeam Infrastructure Appliance with the Veeam Hardened Repository role (recommended) | Certificate-based only; SSH cannot be used | Block storage attached in the Host Management console; disable its Web UI; cannot join a domain; initialise the default Security Officer account first if that account type is enabled; needs hardware RAID and UEFI Secure Boot; no third-party software |
| Manually built Linux server | Single-use SSH credentials for a non-root account with a home directory; remove it from sudoers afterwards | XFS recommended; block storage only (no NFS or SMB mounts); backup folder owned by that account, permissions 0700, no sticky bit; Veeam Agent for Linux must not be installed; physical server with local storage recommended |
| Linux-based backup server (Veeam Software Appliance) | Not applicable | Immutable with no extra setup, but more services mean a larger attack surface; Veeam recommends a dedicated hardened repository for medium and large sites |
A hardened repository cannot be shared between backup servers, does not support symlinks in its path and is not shown in the Files view. An appliance hardened repository can also act as an Application Backup Repository, but the documentation warns this widens the attack surface on immutability.
Common trap: Planning to host a mount server, guest interaction proxy or WAN accelerator on the hardened repository - the only extra role allowed is a Network-mode VMware proxy, and even that is discouraged.
Use only forward incremental with active or synthetic full. Immutable files cannot be merged until they expire, so reverse incremental and forever forward incremental cannot be selected. In v13 reverse incremental is deprecated anyway; it still runs only for jobs that used it before the upgrade. VBM metadata files cannot be immutable (they change every run), so import from VBK files. A backup copy job aimed at a hardened repository needs GFS before immutability applies.
The Veeam Data Mover Service (veeamtransport, port 6162, non-root) receives the period from the backup server. The Veeam Immutability Service (veeamimmureposvc, a root child process) sets and removes the attribute and checks files every 20 minutes. Each backup file gets a .veeam.N.lock file and an extended attribute.
Set-VBRImmutabilityLockExpirationDate.Timeshift detection writes /etc/veeam/immureposvc/timeLog every 10 minutes. If the system or hardware clock shifts by more than 86,400 seconds (24 hours), or the repository is off for more than 24 hours, retention is blocked and the session shows a warning. On an appliance, clear it with Reset time shift protection in the Veeam Host Management console TUI; on a manual server, the fix needs root access. Set the hardware clock to UTC for accurate detection. The documentation states this design protects files even if the NTP server is compromised.
Common trap: Keeping a forever forward incremental job when retargeting it at a hardened repository - immutable files cannot be merged, so only forward incremental with scheduled fulls is allowed.
In v12 many hardened repositories were built from the Veeam Hardened Repository ISO. In v13 they are upgraded in place with the Veeam Infrastructure Appliance ISO. Keep the same hostname and network settings, then switch the managed server to certificate-based authentication. Data stays, but paths move from /mnt/veeam-repository01/ to /var/lib/veeam/. In 13.1, with four-eyes authorisation enabled, reducing or disabling a hardened (or governance-mode) repository’s immutability needs a second approver. Configuration backups can also be stored immutably there (see Chapter 4: Securing the backup infrastructure). A plain Linux repository can be converted by re-adding the server under a different name or IP as a hardened repository and importing the backups. This is not possible for an Infrastructure Appliance repository, a mount server, or a server with packages such as NfsGateway or GuestInteractionProxy installed.
This note is one section of Ultra Transcenders VBR-13: Veeam Backup & Replication 13, an independent study guide that explains every topic the course covers by technology, with comparison tables, diagrams and the common traps, plus a glossary linked to the Veeam Help Center.
Due on Amazon in January 2027, in Kindle and paperback editions.
About the book · VBR-13 terms in the glossary · All VBR-13 study notes
How a backup proxy reads VM data, the order Veeam tries the modes in, and when it falls back to Network mode.
Where the full backup sits in each chain, what each method costs in storage and I/O, and why reverse incremental is deprecated in v13.
How each backup copy mode picks restore points, when it runs and how GFS fulls are created.
Which Veeam technology meets a recovery point and recovery time objective, and what each keeps.