How do I make regular backups under TrueNAS or Unraid?
Which tools or plug-ins can I use to easily back up my data array to an extreme USB hard drive using Unraid or TrueNAS?
I've done a ton of Google research, but I haven't found a truly satisfactory solution anywhere. Rsync is often mentioned, but that only synchronizes my data. However, I would like to have classic incremental backups with version history, like in common Windows programs (Acronis, etc.).
I haven't installed either operating system yet. Currently, my file server is still running Windows 10. But I want to switch and am now looking for the best system. Unraid and TrueNAS offer almost the same capabilities for my application. I'm just not really convinced by the backup aspect of either…
(2 votes)
Loading...
I’m sorry, that should be a reaction to your comment. It was not intended to make a new answer. I clicked there.
These snapshots also read to export to another drive, yes. However, these could also be created much more frequently – because on the one hand it goes super fast, and extra space does not need – only if changed files were covered.
A hard disk array can also be automatically brought back to stand-by. This can also automatically, quite all mechanical plates allow to enter into a time-delayed stand-by mode by inactivity. A read or write attempt on these plates automatically drives them up again, like you know from notebook hard drives. This property has estimated 99.9% of all mechanical hard drives.
I’ll let my snapshots go every hour. Duration to create a snapshot of the complete file system: about 10 milliseconds.
Then I’ll run over a script that thins these snapshots again – i.e. from the hourly snapshots one takes over into the daily snapshots, again one into the weekly, monthly, annual. Because to hold snapshots in 365*24, I’d be a lot. That’s how I’m around and under 50 snapshots.
However, my setup will differ from your intended in some essential details.
No problem has happened to me more often ^^
I’ve read some videos and forum entries for the snapshots now, but somehow don’t really understand. From what I’ve seen now, this function is extremely powerful, but also fucking complicated to understand for a layman.
It also scares me that in most forums it is always recommended to the beginners and instead to simple rsync reflections of the data as backup solution. Because the snapshots seem to make something wrong and the data is gone or damaged…
Oh, man, that’s so tedious! These NAS/Fileserver operating systems are now so powerful and easy to understand. Why could not just be integrated into a completely banal and simple backup function. That would be enough for 90% of non-per users.
I write 10Gigabyte of data on the server every month, so I really don’t need any hourly snapshots…
Unfortunately, the array of backup disks is managed by the malicious RAID controller in the hard disk housing (Lian Li EX-503) and the computer has no direct access to the individual disks. The controller would have the function to shut down the plates, but unfortunately does not work in the practice. I don’t know why.-“
“Data away or damaged” – I put all snapshots as read-only. The data then accidentally harmed excludes that quite. To thin out, I need to make each individual snapshots to be removed (let) writeable again by means of a special command before the one can be deleted.
I actually have more control over the individual drives from RAID because of using md support (quasi “softraid”). I have replaced my previous dedicated NAS/RAID some time ago against an 11x11cm micro pc board and a built-in solution for greater flexibility.
Thank you for star!
Also with rsync can create incremental backups, with some script help
One method is to represent each increment as hardlinks in a separate directory.
Whatever is done is to use snapshots from zfs or btrfs for it.
Is there really no standard solution?
I don’t know about Linux at all and don’t want to make any scripts…. This with the snapshots is also not really great for me and a simple recovery with a complete failure of the array is also not possible without further delay.
This is a file server operating system, why is a simple backup solution not a standard function?!
Recovering snapshots is children’s play – which are shown as directories.
I already find rsync a standard solution. Smooth everywhere, you can do a lot with it, flexible, compact, free. The fact that everyone doesn’t know about it does not stop their usefulness.
Then I get here where the other answer should have come, just the setup here: a replicator, always on, is used by all devices to replicate there. So just sync, no incremental backup. The backup, in turn, is a replication of this replicator, from which snapshots are made.
The replicator can signal the devices to be backupped, but please sync, and otherwise the devices to be backup can sync from themselves with the replicator. Mobile devices and only temporarily available devices preferably make the latter.
This is a remnant of the time when a NAS has been periodically raised for the purpose of backup – which is no longer the case. Another device with mechanical discarray is now constantly online (but mostly idle, device and plates)
This also increases the redundancy.
Data exchange between all devices, including mobile devices, runs all over rsync.
I’m not saying that rsync is not useful either. Not at all, it seems to me to be a very good tool. But still not really designed for my purposes. And with scripts and detours for my project to “use” I find something too far to look for.
Can I also make snapshots monthly on an extrenes drive? The best way to turn off the drive when not in use? (Is an array of 5 hard drives that otherwise consume unnecessary power all the time)