Backup guide

This page provides guidance on creating backups for different level of robustness.

Please note that this page is written to the best of the author's or contributors' knowledge, who may not be data recovery professionals, thus please only treat this as a general guidance.

All commands can be run as normal users.

Casual backup

Below is a simple encoding done using osbx

$ osbx encode file dest_dir

This will yield dest_dir/file.sbx

Sbx containers uses CRC-CCITT for each sbx block.

By default osbx uses sha256 checksum for the entire file content.

The simple encoding allows survival of fragmentation (see official SeqBox README for more information), and typical error detection, but no error correction.

You can duplicate the sbx blocks by simply copying and pasting the sbx container multiple times to lower the chance of missing a valid sbx block for a particular position. This allows higher chance of recovery in case of data corruption.

Adding error correction via Parchive2

You can first encode the file in par2, and then encode each of the par2 files into sbx containers.

We will use 10% redundancy for par2 (-r10) in the following example.

$ par2 create -r10 file

Then we will encode each of the file using osbx

$ for f in file.*par2; do osbx encode $f dest_dir; done

The encoded files are in dest_dir.

Again you may duplicate the sbx containers to increase the chance of finding usable sbx blocks in case of disaster.

Adding encryption

Osbx does not provide encryption of any sort.

Use appropriate tools to encrypt your data before sbx encoding.

Warning

Osbx stores the original file name in the sbx container metadata block. Rename your encrypted file before encoding if your file name may leak information.

We will use gpg symmetric encryption (-c) in the example. Gpg should prompt you for a password.

$ gpg -c file

This yields file.gpg (or file.asc if you added --armour flag).

$ osbx encode file.gpg dest_dir

This yields dest_dir/file.gpg.sbx