I know that for data storage the best bet is a NAS and RAID1 or something in that vein, but what about all the docker containers you are running, carefully configured services on your rpi, installed *arr services on your PC, etc.?
Do you have a simple way to automate backups and re-installs of these as well or are you just resigned to having to eventually reconfigure them all when the SD card fails, your OS needs a reinstall or the disk dies?
There’s lots of very good approaches in the comments.
But I’d like to play the devil’s advocate: how many of you have actually recovered from a disaster that way? Ideally as a test, of course.
A backup system that has never done a restore operations must be assumed to be broken. similar logic should be applied to disaster recovery.
And no: I use Ansible/Docker combined approach that I’m reasonably sure could quite easily recover most stuff, but I’ve not yet fully rebuilt from just that yet.
I restored from a backup when I swapped to a bigger SSD. Worked perfectly first try. I use rsnapshot for backups.
I’m not sure what Ansible does that a simple Docker Compose doesn’t yet but I will look into it more!
My real backup test run will be soon I think - for now I’m moving from windows to docker, but eventually I want to get an older laptop, put linux on it and just move everything to the docker on it instead and pretend it’s a server. The less “critical” stuff I have on my main PC, the less I’m going to cry when I inevitably have to reinstall the OS or replace the drives.
I just use Ansible to prepare the OS, set up a dedicated user, install/setup Rootless Docker and then Sync all the docker compose files from the same repo to the appropriate server and launch/update as necessary. I also use it to centrally administer any cron jobs like for backup.
Basically if I didn’t forget anything (which is always possible) I should be able to pick a brand new RPi with an SSD and replace one of mine with a single command.
It also allows me to keep my entire setup “documented” and configured in a single git repository.
While rsync is great, I recovered partially from an outtage… Containers with databases need special care: dumping there database…
Lesson learned !
I’ve had a complete drive failure twice within the last year (really old hardware) and my ansible + docker + backup made it really easy to recover from. I got new hardware and was back up and running within a few hours.
All of your services setup should be automated (through docker-compose or ansible or whatever) and all your configuration data should be backed up. This should make it easy to migrate services from one machine to another, and also to recover from a disaster.
Pre…pare…? What’s that? Some sorta fruit?
My configuration and deployment is managed entirely via an Ansible playbook repository. In case of absolute disaster, I just have to redeploy the playbook. I do run all my stuff on top of mirrored drives so a single failure isn’t disastrous if I replace the drive quickly enough.
For when that’s not enough, the data itself is backed up hourly (via ZFS snapshots) to a spare pair of drives and nightly to S3 buckets in the cloud (via restic). Everything automated with systemd timers and some scripts. The configuration for these backups is part of the playbooks of course. I test the backups every 6 months by trying to reproduce all the services in a test VM. This has identified issues with my restoration procedure (mostly due to potential UID mismatches).
And yes, I have once been forced to reinstall from scratch and I managed to do that rather quickly through a combination of playbooks and well tested backups.
Dang I really like your idea of testing the backup in a VM… I was worried about how I’d test mine since I only have the one machine, but a VM on my desktop or something should do just fine.
I rsync my root and everything under it to a NAS, will hopefully save my data. I wrote some scripts manually to do that.
I think the next best thing to do is to doco your setup as mich as possible. Either by typed up notes, or ansible/packer/whatever, any documentation is better than nothing if you have to rebuild.
I have a 16tb USB HDD that syncs to my NAS whenever my workstation is idle for 20 minutes.
I run history and then clean it up so i have a guide to follow on the next setup. It’s not even so much for drive failure but to move to the newer OS versions when available.
The ‘data’ is backed up by scripts that tar folders up and scp them off to another server.
Radical suggestion:
- Once a year you buy a hard drive that can handle all of your data.
- rsync everything to it
- unplug it, put it back in cold storage
- Back everything up
rm -rf /- Now rebuild.
Congratulations, you now know what’s required. :-P
Rebuild to different disks than the ones you backed up though. Don’t restore over your working data.
Infrastructure as code/config as code.
The configurations of all the actual machines is managed by Puppet, with all its configs in a git repo. All the actual applications are deployed on top of Kubernetes, with all the configurations managed by helmfile and also tracked in git. I don’t set anything up - I describe how I want things configured, and the tools do the actual work.
There is a “cold start” issue in my scheme - puppet requires a server component that runs on Kubernetes but I can’t deploy onto kubernetes until the host machines have had their puppet manifests applied, but at that point I can just read the code and do enough of the config by hand to bootstrap everything up from scratch if I have to
I have all my configuration as Ansible and Terraform code, so everything can be destroyed and recreated with no effort.
When it comes to the data, I made some bash script to copy, compress, encrypt and upload them encrypted. Not sure if this is the best but it is how I’m dealing with it right now.
I’ve got a similar setup, but use Kopia for backup which does all that you describe but also handles deduplication of data very well.
For example I’ve added older less structured backups to my “good” backup now and since there is a lot of duplication between a 4 year old backup and a 5 year old backup it barely increased the storage space usage.
That sounds a lot like how I keep my stuff safe, I use backblaze for my off-site backup
By using NixOS and tracking the config files with git
For real, saves so much space that would be used for VM backups.
Aside from that, I have anything important backed up to my NAS, and Duplicati backs up from there to Backblaze B2.
- Install Debian stable with the ssh server included.
- Keep a list of the packages that were installed after (there aren’t many but still).
- All docker containers have their compose files and persistent user data on a RAID1 array.
- Have a backup running that rsyncs once a day /etc, /home/user and /mnt/array1/docker to another RAID1 to daily/, from daily/ once a week rsync to weekly/, from weekly/ once a monthb timestamped tarball to monthly/. Once a month I also bring out a HDD from the drawer and do a backup of monthly/ with Borg.
For recovery:
- Reinstall Debian + extra packages.
- Restore the docker compose and persistent files.
- Run docker compose on containers.
Note that some data may need additional handling, for example databases should be dumped not rsunced.
My docker containers are all configured via docker compose so I just tar the .yml files and the outside data volumes and backup that to an external drive.
For configs living in /etc you can also backup all of them but I guess its harder to remember what you modified and where so this is why you document your setup step by step.
Something nice and easy I use for personal documentations is mdbooks.
Ahh, so the best docker practice is to always just use outside data volumes and backup those separately, seems kinda obvious in retrospect. What about mounting them directly to the NAS (or even running docker from NAS?), for local networks the performance is probably good enough? That way I wouldn’t have to schedule regular syncs and transfers between “local” device storage and NAS? Dunno if it would have a negative effect on drive longevity compared to just running a daily backup.
If you’ve got a good network path NFS mounts work great. Don’t forget to also back up your compose files. Then bringing a machine back up is just a case of running them.
I actually run everything in VMs and have two hypervisors that sync everything to each other constantly, so I have hot failover capability. They also back up their live VMs to each other every day or week depending on the criticality of the VM. That way I also have some protection against OS issues or a wonky update.
Probably overkill for a self hosted setup but I’d rather spend money than time fixing shit because I’m lazy.
HA is not redundancy. It may protect from a drive failure but it completely ignores data corruption issues.
I learned this the hard way when my cryptomator decided to corrupt some of my files, and I noticed but didn’t have backups.
That’s why I also do backups, as I mentioned.
yeah, there’s a bunch of lessons that tend to only be learned the hard way, despite most guides mentioning them.
similarly to how RAID should not be treated as a backup.
My server has a raid1 mdadm boot drive. And an 8 dive raid6 with zfs. It’s been running for 14 years now. The only thing that I haven’t replaced over it’s lifetime is the chassis. In fact the proc let out the magic smoke a few weeks ago, after some new parts it’s still going strong.
I put all docker data in one directory (or rather, a btrfs subvolume) and both snapshot and back it up daily to multiple machines.
docker-composefiles are also kept in the same subvolume.My latest server is NixOS, so I don’t even bother backing up the root subvolume, since the actual config is tracked on git and replicated on multiple machines. If I want to reinstall, I can just install NixOS and deploy the config, then just copy over the docker subvolume, and rebuild the containers. Some of this could be automated further (
nixos-anywhereanddiskolook promising for the actual OS install) but my systems don’t typically break often enough for that to be a significant issue.You can go even further and either just use nix for the services, or use nix to build containers themselves, but I have a working setup already and it’s good enough, and I can easily switch to another distribution if issues start occurring in NixOS.
