Home » CloudZat UK » Borg Backup: Create an Encrypted Archive and Test Recovery

Borg Backup: Create an Encrypted Archive and Test Recovery

CLOUD BACKUP & PHOTO LIBRARIES

Borg earns its place when you need versioned file backups without uploading unchanged content repeatedly. The important first milestone is an archive you can extract, with its key available away from the original machine.

Editorial review: 15 November 2026 · UK

Quick answer

Initialize an encrypted repository, create a named archive, list its contents and extract a sample into a separate directory before scheduling backups.

Steps to follow

  1. Before you start

    Keep the original data available while following this procedure. Record the application version, account, source path and destination you are actually using. If an error is involved, save its exact wording. Use a small, non-sensitive sample first so that an unexpected result is easy to inspect without disturbing your working library.

  2. Set up the right scope

    Check borg –version and use the matching manual. For Borg 1.x, initialize a new destination with borg init –encryption=repokey /backup/repo. Choose your own destination and retain the passphrase and exported key securely.

  3. Apply the relevant settings

    Create a uniquely named archive with borg create /backup/repo::first /path/to/data. Check warnings about unreadable or changing files. Use consistent database exports rather than assuming a live database directory is a valid backup.

  4. Complete the first run

    Use borg list /backup/repo and inspect the selected archive. Change into an empty recovery directory before borg extract /backup/repo::first; extraction writes relative to the working directory.

Decisions and checks

The limitation that matters

Borg 1.x and 2.x commands and repository formats differ. Do not mix examples across versions. Keep free space for repository operations and review retention before enabling prune or compact.

How to verify the result

Open recovered documents and compare several sizes or hashes. Test recovery using the recorded key and passphrase, then schedule runs with logs and failure notifications.

Keep the result reliable

After the first successful check, repeat the process with an ordinary working session or scheduled run. Record the date and the recovery location. Revisit the setup after changing devices, accounts, storage paths or software versions; yesterday’s successful job does not prove that a new configuration is covered.

Compare the options

Situation / component Purpose / action What to check
Input Preserve an unchanged sample Know which account and location contain the original
Operation Follow the procedure for your installed version Review errors and skipped items, not only the completion message
Result Open the resulting file or test the documented recovery route Confirm that the content, version and access rights are correct

Frequently asked questions

Can I remove the original immediately?

Keep it until you have verified the intended result and have an independent recovery copy where the data matters. A visible file name, a green icon or a completed task alone does not show that every needed item is recoverable.

What information helps if this fails?

Record the version, exact error, affected file type, available space and whether the same small sample works elsewhere. Share sanitized details through the official support route; passwords, recovery keys and private documents are not needed for an initial diagnosis.

Official documentation

This guide combines current official documentation with practical recovery checks. It does not claim hands-on testing of your installation. Menu labels and supported procedures can vary by release; follow the documentation for your installed version.

Scroll to Top