My backup approach (HomeBox on Synology, remote access via Tailscale)

[danielrosehill]

Hello fellow HomeBoxers!

As a long-term Linux and open-source tinkerer ... and sometimes breaker of these things ... I always try to think of "backup-ability" from the get go whenever using a new technology.

I've deployed HomeBox as a container on my Synology NAS. I followed the instructions here to do so.

This is what I'm currently doing for backup (I shared these ideas a video on YouTube but figured I'd duplicate here in text). Not suggesting for a moment that this is either the best or one and only approach. But it's fairly simple and gets the job done.

1 - Store user data on a separate and dedicated volume

The first thing I did was move my user data over to a separate volume on the NAS.

For lack of imagination, I called mine "HomeBoxUserData".

"Transplanting" my user data from the original volume (nested in /docker) over to a separate volume worked perfectly (to be safe, I powered down the container while migrating the data).

At the time of writing, the HomeBox user data folder contains:

• Three database files at the root level
• One subdirectory in which all the "attachments" are stored (photos appended to items etc).

You can backup the container itself (in Synology it's very easy. Action -> Export from Container Manager). But doing so makes little sense as the container will be modified after every system update. So the user data volume is really the target.

For onsite backups, I'm using Snapshot Replication (DSM Package)

Some backup purists might wince at the idea of taking snapshots on the same physical machine as the target but ... hey, I'm living on the wild side here!

Snapshot Replication is a nice tool and allows you to create as many snapshots as you like. This makes rolling back unwanted changes very easy (at the click of a button, literally).

Snapshot Replication works at the volume level which is why I moved my user data over to a volume.

Here's my user data folder with daily snapshots configured (I retain two dailies and take the odd 'on demand' one).

Personally, I like to see the files that the tool is creating. So in 'advanced' I enabled 'make snapshot visible'. We'll see why in the final step:

**3: Set up offsite duplication using Cloud Sync **

The final step in the process is that we need to duplicate our backups/snapshot offsite. There's a replication tool within Snapshot Replication but I find Cloud Sync much easier to work with.

Simply add your preferred cloud storage and then set up a sync job to upload the snapshots. Your source will be the user data volume and your target will be a directory you create in your cloud storage.

I have my offsite push running weekly.

Note: Cloud Sync won't "see" the snapshot directory so you'll only be uploading one copy and not the various point-in-time snapshots.

Other approaches

You can write rsync scripts in Task Scheduler (and use rsnapshot in the Community Packages) in order to achieve roughly the same functionality as I described above. Just ... less GUIs and a bit more legwork to configure.

Hope that was of some help :-)