I was wondering how often does one choose to make and keep back ups. I know that “It depends on your business needs”, but that is rather vague and unsatisfying, so I was hoping to hear some heuristics from the community. Like say I had a workstation/desktop that is acting as a server at a shop (taking inventory / sales receipts) and would be using something like timeshift to keep snapshots. I feel like keeping two daily and a weekly would be alright for a store, since the two most recent would not be too old or something. I also feel like using the hourly snapshots would be too taxing on a CPU and might be using to much disk space.
I still have drawings I made in MS Paint on Windows 95 when it had just come out, my first text document, and the first report I ever typed in grade school.
Btrfs snapshots of the root volume in RAID1 configuration with 8 hourly, 7 daily, 3 weekly, and automated rsync backups to NAS, with primary and secondary offsite, physically disconnected backups stored in sealed, airtight, and waterproof containers at two different banks prepaid storage and with advanced directive in the event of my demise.
Bit of a hobby really. I acknowledge it’s completely unnecessary. I don’t like to lose data.
Sealed, airtight, and waterproof but what if both banks burn down at the same time? You didn’t mention fire-proof.
You got me there! Not fireproof. In that case I’m just hoping that having two off-site backups at different locations has me covered, but that’s a good idea. I should consider fireproof foil.
You are backed up better than some enterprises…
Just wow.
Another perspective is data hoarding.
I have system images of machines of relatives who have died. Many of the photos that I have retained are the only ones. However, that was more an emergent utility than a motivating one.
How often do you update your off-site backups?
Monthly, alternating locations.
Never
I back up every morning to get to work and every afternoon to get back home lol
How often depends on how much work it is to recreate, or the consequences of loosing data.
Some systems do not have real data locally, get a backup every week. Most get a nightly backup. Some with a high rate of change , get a lunch/middle of the workday run.
Some have hourly backups/snapshots, where recreating data is impossible. CriticL databases have hourly + transaction log streaming offsite.How long to keep a history depends on how likely an error can go unnoticed but minimum 14 days. Most have 10 dailes + 5 weeky + 6 monthly + 1 yearly.
If you have paper recipes and can recreate data lost easily. Daily seems fine.
My PC: Every day and when it is online
My drives the backups go to: Once a week.Wait, so you backup your backups? Why not just 2 backups of the same thing?
Limitation of hardware.
It is essentially just a file copy.
As others have said, with an incremental filesystem level mechanism, the backup process won’t be too taxing for the CPU. I have ZFS set up which makes this easy and I make hourly snapshots using sanoid which also get sent to another mirrored pair of connected drives using syncoid. Then, once a day, I upload encrypted daily snapshots to a bucket in the cloud using restic. Sounds complicated, but actually sanoid/syncoid and restic do all the heavy lifting. All I did is automate their schedules using systemd timers and some scripts to backup the right directories.
I upload encrypted daily snapshots to a bucket in the cloud using restic.
How do you upload a snapshot? I’m using TrueNAS where I can make snapshots visible in a otherwise hidden .zfs directory. Do you just backup from there or something similar? Is there an upside to backing up a snapshot instead of just the current data?
How do you upload a snapshot?
Basically, as you said. Mount the data somewhere and back up its contents.
I back up snapshots rather than current data, because I don’t want to stop the running containers that read and write from that data. I’d rather avoid the situation where the container is writing data while it’s being backed up. The back up happens shortly after the daily snapshot is made so the difference between current and snapshot data is small.
I’m probably the overkill case because I have AD+vC and a ton of VMs.
RPO 24H for main desktop and critical VMs like vCenter, domain controllers, DHCP, DNS, Unifi controller, etc.
Twice a week for laptops and remote desktop target VMs
Once a week for everything else.
Backups are kept: (may be plus or minus a bit)
- Daily backups for a week
- Weekly backups for a month
- Monthly backups for a year
- Yearly backups for 2-3y
The software I have (Synology Active Backup) captures data using incremental backups where possible, but if it loses its incremental marker (system restore in windows, change-block tracking in VMware, rsync for file servers), it will generate a full backup and deduplicate (iirc).
From the many times this has saved me from various bad things happening for various reasons, I want to say the RTO is about 2-6h for a VM to restore and 18 for a desktop to restore from the point at which I decide to go back to a backup.
Right now my main limitation is my poor quad core Synology is running a little hot on the CPU front, so some of those have farther apart RPOs than I’d like.
Daily, usually keeping only the last week or so
There should be a whitepaper you can reference based on sales scenario. As others have said hourly, daily, weekly snapshots are not backups, unless you also have a btrfs or zfs send that IS backing up the snapshots to another remote device
Like you said, “it depends” 😁
I have a huge datablob that I mirror off-site once monthly. I have a few services that provides things for my family, I take a backup of them nightly (and run a “backup-restoration” scenario every six months). For my desktop, none at all - but I have my most critical data synched / documented so they can be restored to a functional state.
When you use deduplication on the backup side you can do backups every minute without needing much storage. When the backup programm looks at the filesystem to determine which file has changed, the CPU only need to process the changed files.
For my personal devices i do daily backups. There is not enough change every day.
Whenever I’ve gone too far forwards
IMO set up a good incremental backup system with deduplication and then back up everything at least once a day as a baseline. Anything that’s especially valuable can be backed up more frequently, but the price/effort of backing up at least once a day should become trivial if everything is set up correctly.
If you feel like hourly snapshots would be worth it, but too resource-intensive, then maybe replacing them with local snapshots of the file system (which are basically free, if your OS/filesystem supports them) might be reasonable. Those obviously don’t protect against hardware failure, but help against accidental deletion.
I have autocron jobs that sync various server directories to a daily backup (on the same server), then sync that backup once a week to the weekly backup, and once a month take a tarball snapshot of the weekly backup.
Every once in a while I plug in a HDD on USB and take a Borg backup of the monthly dir. Borg does compression and deduplication (and encryption if you want to). I should be doing this also once a week but sometimes I’m lazy and leave a few weeks between them.
Snapahots are not backups!
Snapshots are near instand in ZFS or BTRFS. Also they do not consome much CPU to make them.
Backups are not stored in the same device. What i use to make backups is a combination of borg and urbackup.
It depends what I’m backing up and where it’s backing up to.
I do local/lan backups at a much higher rate because there’s more bandwidth to spare and effectively free storage. So for those as often as every 10 mins if there are changes to back up.
For less critical things and/or cloud backups I have a less frequent schedule as losing more time on those is less critical and it costs more to store on the cloud.
I use Kopia for backups on all my servers and desktop/laptop.
I’ve been very happy with it, it’s FOSS and it saved my ass when Windows Update corrupted my bitlocker disk and I lost everything. That was also the last straw that put me on Linux full-time.