After recently upgrading my main storage server bit by bit, I found myself with a pile of parts that basically added up to another storage server. I had a capable Network Attached Storage (NAS) solution for files and media, but I needed a second server running on-site as a backup for the main NAS. What I did not want, however, was to pay for a second server running 24/7 gobbling up energy. This article explains how to set up a pair of servers in a primary/backup configuration so that the backup will synchronize itself with the primary server each day, week, month, or however often you like. Once the two are synced, the backup server will turn back off until needed again, thus saving most of the energy costs. This approach has an added benefit in the event of a ransomware encryption attack, because the backup server will most likely be turned off at the time of the attack, making it more likely to escape encryption.

Hardware Configuration

The two servers I describe in this article are used in a home lab and are non-critical. If you are using servers for a business, school, or professional agency, you will have different needs and, ideally, a more expansive budget that could point you toward a different solution. This article is intended as a proof of concept – and as a way to explore some of the tools available in the Linux environment. This basic approach might very well be feasible for other secondary or off-site backups with slight modifications.

The main NAS server has six 6TB HGST SATA drives in a RAIDZ2 pool with approximately 24TB of usable storage space (Figure 1). The pool can lose two drives and still retain all of the data, but obviously doing this means 33 percent of the disks' raw space is unavailable. Being in a RAID array means that there is redundancy, but redundancy is not the same as a backup. Both servers have 10Gb networking, and the primary NAS runs a Proxmox virtual environment [1], which uses SMB to share to the backup.

[...]