Paperback offline backup

For disaster recovery purposes, it is crucial to always have access to all secrets that are needed to black start your environments from zero. Depending on your setup, those credentials can be

• root passwords for secret managers
• passwords for encrypted backups
• GPG encryption keys
• root credentials for cloud accounts
• private keys for PKI infrastructures

To make sure the backup is secure and reliable, the following requirements should be met

• it should be human and machine-readable, in a clear and easy-to-read format that is trivial to convert back to digital
• should not be stored in plaintext but use a proper encryption algorithm
• the encryption should support distributed trust, so that no single person in possession of the backup can access the secrets alone
• be stored on non-digital mediums to avoid accidental deletion or exposure
• have multiple copies that are geographically redundant and distributed to protect from single-point of failure (e.g., house fire, flood)
• stored in a secure, tamper-proof location
• written guides for rotation and recovery exist and are regularly tested and trained
• include checksums or validation words to be able to verify the correct entry during recovery

paperback

paperback is a paper-based backup scheme that is secure, easy-to-use and can fulfil those requirements. Backups are encrypted, and the secret key is split into numerous key shards which can be stored separately (by different individuals), removing the need for any individual to memorise a secret passphrase.

Download

Backup

To create a backup of sensitive data, execute the following steps

Write secrets to a file

backup.txt

foo-bar

Create backup documents from file

Depending on your needs, define the total number of key_shards you want to distribute and a quorum which is the number of key shards needed to restore the secrets from the backup

paperback backup --quorum-size <quorum> --shards <key_shards> backup.txt

Distribute PDFs

paperback backup will create a main document with a unique document_id and n key shards with unique key_shard_id s

main_document-<document_id>.pdf
key_shard-<document_id>-<key_shard_id>.pdf
key_shard-<document_id>-<key_shard_id>.pdf
key_shard-<document_id>-<key_shard_id>.pdf
key_shard-<document_id>-<key_shard_id>.pdf

Each storage location should contain a copy of the main document and a single key shard.

Main document

The main document consists of

• ❶ the document_id
• ❷ the encrypted document data
• ❸ a checksum that can be used to verify document data entry during restore

Key shard

The key shard document contains

• ❹ the key_shard_id
• ❺ the document_id
• ❻ the encrypted key shard data
• ❼ a checksum that can be used to verify the key shard data entry during restore
• ❽ codewords to decrypt the key shard data

Clean up

After document distribution, clean up the machine where the backup was created and remove the PDFs and the secret file.

Recovery

Start the interactive recovery process with the target file where the record data should be written to

paperback recover --interactive recover.txt

To recover, you need to enter

• the encrypted document data (❷)
• a <quorum> number of
  • encrypted key shard data (❻)
  • codewords to decrypt the key shard data (❽)

into the recovery process

Enter a main document code (unknown number of codes remaining): <❷>

Main document checksum: ❸
Document ID: <document_id>
<quorum> key shards required.
Quorum contains [] key shards.
Enter key shard 1 of 2: <❻>

Key shard 1 checksum: ❼
Enter key shard 1 codewords: <❽>

Loaded key shard ❹.
Quorum contains [❹] key shards.
Enter key shard 2 of 2: <❻>

Key shard 2 checksum: ❼
Enter key shard 2 codewords: <❽>

Loaded key shard ❹.

The recovered data will be written to recover.txt