By Backrest

Backrest

What it does

Backrest adds a browser-based control layer to restic. You connect or create a repository, choose source paths and a schedule, then use its operation history and snapshot browser to inspect runs and restore files. Restic remains the backup engine.

It is a self-managed setup: the repository can be local or use a supported remote storage provider, but the storage account, credentials and host operation remain your responsibility. Backrest is most useful when you want restic backups organized through plans and visible operations.

Last updated
Oct 27, 2026
Backrest · Editorial illustration
Editorial illustration of source folders moving through a scheduled backup into stored snapshots, with a restore path. Not a product screenshot.

Work at Backrest?

Help customers find accurate information about your software.

Manage this listing →

Find your fit

Who it is for

A small team or technically comfortable operator already using restic can connect an existing repository, index earlier snapshots and add scheduled plans without discarding the underlying backup format.

Someone responsible for several machines can consider paired Backrest instances to collect operation history and share selected configuration. Review the permissions and network exposure before using a central server.

A buyer seeking a fully hosted backup service should budget for the separate storage provider and administration of the Backrest host. The official setup assumes those pieces are supplied by the operator.

Back to sections

The work it supports

Core capabilities

Add a new or existing restic repository, enter its storage location and encryption credentials, then index earlier snapshots if needed. Repository settings can include maintenance schedules.

Build a backup plan by choosing paths, exclusions, repository and a schedule. Interval or cron timing can use local, UTC or last-run clocks, so test a harmless run to confirm the host executes when expected.

Set a count-based or time-based retention policy if wanted. Forget marks snapshots outside that policy, while prune removes unreferenced data; review both before depending on a particular recovery window.

Schedule repository integrity checks and choose how much data to read. A full data read can be slow and can trigger remote-provider egress costs; a separate restore test is still needed to confirm useful recovery.

Browse indexed snapshots, choose a file or directory, and restore it to a selected path. Backrest shows the restore as an operation; remote repositories can make browsing slower because restic must fetch data.

Inspect operation history and diagnostic logs when a run fails. Hooks can send configured notifications or execute commands on success, error and other lifecycle events.

Paired instances can share selected repository configuration and sync operation history for one view across hosts. Pairing permissions can be scoped; keep administrative UI endpoints on a trusted network when allowing remote sync.

Save repository credentials, encryption keys and Backrest configuration securely outside the machine being protected. The documented restore path depends on those credentials being available.

Back to sections

Understand the offer

Pricing and getting started

The documentation describes installing and running Backrest yourself with your own local or remote repository. A remote storage provider, credentials and any storage or transfer charges are separate; check those terms for the destination you choose.

Backrest runs operations through restic and can obtain a compatible binary when one is not already present. Account for that dependency and your host’s installation policy when planning deployment.

Back to sections

Look a little closer

Before you choose

Try one non-sensitive backup plan end to end: select source paths, run or wait for a scheduled snapshot, inspect the operation result, and restore a sample file to a different directory. That checks the recovery path more directly than a green job status.

Check how long snapshots should remain and whether forget, prune and integrity-check schedules match the available storage and provider charges. Keep enough recovery points for the kinds of mistakes you need to undo.

If central monitoring matters, pair test instances with narrow permissions and confirm that the admin UI is reachable only where intended. Preserve repository keys and configuration outside the protected host.

Back to sections

6 distinct sources behind this profile. Research observed Oct 27, 2026. Features and offers can change.
View sources and research →

Help keep this profile useful

Spotted an outdated detail? Send the correction with a supporting source. A verified account is needed to submit.

Suggest a correction ↗