Hi all,

I have a large personal photo/video collection built up over the years and want to make sure it’s properly protected. Right now everything sits on two external HDDs, and I’ve read that drives have a limited shelf life, so I’m looking to build a more reliable long-term system.

A few things I’d love input on:

The 3-2-1 backup rule: I understand the basic idea (3 copies, 2 different media types, 1 offsite), but how do people actually implement this in practice for a home photo/video archive? What’s your setup, brands, models, and how it’s all connected?

File formats: should I be converting anything, or keeping the original format?

Backup methods: NAS vs. cloud vs. cold storage HDDs vs. optical media (M-DISC, etc.)? What’s actually held up well for you over 5-10+ years?

HDD lifespan: how often do you actually need to replace/refresh drives, and how do you monitor drive health before something fails?

Specifics/cost: if you’re using a NAS, which model, how many bays, and what did it cost to set up? If cloud, which provider and what’s the monthly/yearly cost for your storage amount? If cold storage HDDs, which brand/model do you trust, and roughly how much per TB?

Would really appreciate hearing what’s worked (and what hasn’t) for people who’ve been doing this a while, including rough costs so I can budget properly.

I used to be one of those people who just copy-pasted folders onto an external drive and that was it. Now I want to do something a bit more elaborate.

I had heard of Clonezilla and learned to use it at a basic level, creating clones, disk images, and performing restores. However, as a system it has a lot of downsides:

• It only does full backups, not incremental ones; if 10 MB of important information changes, you have to back up the whole disk from scratch.
• You have to shut down the PC and use a USB stick, which is impractical for frequent backups.
• It's not portable to other machines! A disaster recovery scenario I don't see mentioned is if not only the disk but the whole PC dies, and you buy a new one with different specs. In that case, the Clonezilla clone or image cannot be restored onto the new machine.

So I looked into it further and found other programs like Borg and Restic. Haven't tried them yet, but they seem useful at the file level.

How do you all handle this? Do you use a single backup program or a combination of them?

Always considering my particular use case (home use, Linux, no cloud, and only external drives).

I used cloudberry to back up my files on an old Ubuntu computer about 10 years ago. I'm trying to get the files back of AWS but none of them will open. When I open some text files they are all in Chinese. I'll include an example in the comments if I can.

Is it possible to get Acronis 2025 with the perpetual license for offline use? I'm looking to put it on a thumb drive to download and use on a pc complete offline.

We’re backing up roughly 900TB on prem right now. Environment is a bit of everything: vSphere VMs, SQL and Oracle, some physical servers, AD, and NAS. We’re on Rubrik at the moment, but the cost is getting harder to justify, so I’ve been looking at what else is out there.

One complication is that we probably won’t be on VMware forever. OpenShift is looking like the more likely direction, so I don’t really want to move to another backup platform that works well for VMware today but leaves us stuck again in a few years. Main requirements are reliable recovery for VMs and databases, physical server support, NAS, and ideally decent support for OpenShift/container workloads as that side grows.

Anyone here managing a similar-sized mixed environment? What did you end up using, and what has the day-to-day experience been like after the migration?

I fear my laptop may have been compromised. I stupidly never bought an SSD before the prices have literally quadrupled. I can't justify a purchase now.

All my photos are on my laptop. I want to reformat my laptop so I can use it again but I don't want to lose years worth of photos.

Is backing up to the cloud reformatting then re-downloading my only option other than buying an HDD or SSD??

What i can do? I store in my personal nas, but ita geting bigger and bigger, i just deleted all? Store in better way? Online in freeze bucket? Seens to be too much effort, please help

Following up on my post from four months ago, I went on the hunt for a Dropbox backup application. CloudAlly checked all the boxes, did an overview with a sales rep, got the trial going, did a few test restores, and all was well. Three months later I'm looking at my quarterly vendor spending report, and I see that we're paying over $350.00 a month for a mere 2TB of data for this one client. For that price, we could be self hosting a backup with versioning and retension on a Synology NAS, though that comes with its own risks both through technical and legal exposure.

Other than self-hosting, what other vendors should I be looking at?

Hi all

I need a sense check on my backup/recovery system as my family glaze over if i talk to them about it . i have a few questions in relation to this

The Process
All our devices - windows laptops/ phones / tablets sync with my Synology NAS - The NAS is encrypted

The important bits of the NAS are backed up weekly via synology hyper drive - -and encrypted on 2 USB sticks - one of which is off site and rotated - semi regularly

Each person uses Proton Pass for their passwords end Ente Auth for 2fa codes.

The recovery keys for NAS encryption / ente / proton pass are saved in proton pass (shared with family) , copy on the nas / copy in my proton drive

There is also a seperate copy of the recovery keys on the rotating USB - encrypted with cryptomater - this password is in proton pass but this is the one password i have memorised

Questions i could do with advice on

1 - is it too much/ too complicated

2 - do i also need a cloud backup of NAS

3 - can i simplify it

4 - do i have enough redundancy

Thanks i advance for your help

Hi as the title already says, do I want to ask what the best way is to save my whole
CachyOS-system on an external drive. No fancy automatic saving on a NAS or something like that, just simply plugging the drive in to the PC making with for example a program a whole save on my drive and that's it. And can also someone explain that if it should happen that something happens to the PC how I can revive my system with the save?

Hello guys, I have an Immich server with a total size of around 45 GB, with the database, uploads, and everything else stored under a single directory. I want to back it up to Backblaze using Restic. How should I go about doing it?

I have some idea of how to do it. I was thinking about selectively backing up the upload directory and database dumps instead of the entire directory. Would that be a better approach, or should I just let Restic back up the entire Immich directory?

I also have fairly limited upload bandwidth (around 5 Mb/s), so I’d like the backup process to be able to run gradually in the background and resume if my PC or connection is restarted.

What do you guys think? How are you running your Immich backups, and is there anything specific I should make sure to back up so that I can fully restore Immich if my PC dies?

Also, I’d love to know about your backup strategies.

Hi all.

I am mainly using this for personal data: old photos, textbooks, school projects, reference data, videos, movies (ahhem, yes) and legacy programs (ahhem, yes) and tax documents (receipts, etc).

Main daily use is my personal laptop (not much inside), just for YouTube, some photo editing and such, document editing and such. My work is now backed to cloud server so that's pretty much ok. Might be a data horder...

Not a techie...

Using a windows os

Around 50GB to 100GB of data*

Personal storage 500GB SDD, 500GB HDD, 250GB HDD

Current system: Manually copy and paste

1. Important stuff gets copied everywhere (30GB).

2. Backups (half yearly, if i remember) this is done incremental (e.g. year 1: 50GB, year two 50GB + 10GB, year three 50GB+10GB+10GB; total 80GB)

3. Its really messy and I take a long time to find my files. Just thought i lost my important backup and found it hidden physically (the sdd) in my office - would like to avoid this too.

So i would like to arrange my datas so that I can access them easier and stored easier.

Budget minimum: i dont really want to spend more than I have to. And judging by the prices of storage... probably i am looking at spending less than usd 500 for an entire system. One in use and one offsite. I intend for this to be offline usage so avoid virus or other issues.

The easier way will be to use more hard drive to continue my Backups and perhaps delete old datas. But I wouldnt know if these data are corrupted during copying - it happened once before.

So please suggest a easy backup and forget system with some form of checking for corruption.

My current Backup setup:

• I have 2 24TB Windows storage spaces compromised of 2X12TB drives each.
  • Storage Pool A consists of my media and archives
  • Storage Pool B is called Backup
    • Files and Folders backup using Macrium 8, 1 full, 5 weekly incremental, followed by delete and full backup of
    • C:/users/user/onedrive
    • Storage Pool A
    • D:/
    • Folder Sync Pro of my S24 Ultra (nightly 1 way sync, do not sync deletions)
    • Gallery (Also backed up to Onedrive and Google Photos)
    • Whatsapp (also backed up to Gdrive)
• My C:/ drive is a 2TB Samsung 990 Pro SSD
  • Contains Windows and my OneDrive (set to all files on device)
• My D:/ Adata  1TB SSD
  • Contains overlfow Pictures and Family Videos
• My X:/ 2TB Adata NvME
  • Weekly clone of C://
• IDrive Cloud backup unlimited
  • Realtime backup of Storage Pool B
  • 30 day versioning

The only hole I see in mysetup is I do not know when bitrot will occur in my storage pools.

I had already run into this once, with an archive from 2010, that last year I discovered was corrupted because a drive failed in Storage Pool A (without errors).

I only knew bitrot had occurred when I then tried a few other archives and they too were corrupted. Even some pictures that were in Google Photos, were showing as an exclamation mark icon in the local file preview. I am glad I had the Google Photos copy, but the zip archives were gone forever. Even the 30 day versioning of IDrive was not useful, as this was discovered 30 days after corruption. Nothing had alerted me to the drive deterioration. It was completely silent. I paid for Stablebit scanner after that, but not sure if that is sufficient, or is it better now to get a NAS and setup RAID?

Thanks

Dropbox Camera Uploads dumps everything into one folder, just renamed with a date in the filename — no automatic subfolders, no location sorting. Same story with most tools I've tried. I'm looking for something that actually creates folders by location + date automatically, and can purge originals from the source device after a set retention period.

Does anything handle this natively, or is it always scripts / manual tagging on top of a flat upload?

Any suggestions on the best cloud storage or cloud backup apps for iOS and Android in 2026 ?

I searched for a good backup tool to back up my linux system (root directory) to an external HDD.
I found recommendations for restic, borg and kopia so here are my personal tests and observations.
This post isn't really for discussion but rather an information dump and I hope that it will be helpful for people in the same situation as me.

First, I did some performance tests.
My root directory lives on an internal SSD and I used borg v1.4.5, restic v0.19.1 and kopia v0.23.1.

Created uncompressed backups of the root folder (bind mounted at /mnt via mount --bind --make-private / /mnt):

Borg: 25.1 GB in 13:55m, using 43 % CPU on average and a maximum of 272 MB RAM.
Restic: 24.1 GB in 14:57m, using 66 % CPU on average and a maximum of 381 MB RAM.
Kopia: 24.6 GB in 10:30m, using 61 % CPU on average and a maximum of 836 MB RAM.

I ran the backup twice to see how fast the rescan time is:

Borg: 1:31m
Restic: 0:34m
Kopia: 0:10m

Cache sizes:

~/.cache/borg: 306 MB
~/.cache/restic: 173 MB
~/.cache/kopia: 1128 MB

Created zstd compressed backups of the root folder (bind mounted at /mnt):

Borg (--compression auto,zstd,3): 10.4 GB in 11:15m, using 59 % CPU on average and a maximum of 262 MB RAM.
Restic (--compression auto (=level 3)): 10.4 GB in 7:44m, using 304 % CPU on average and a maximum of 393 MB RAM.
Kopia (--compression zstd): 10.3 GB in 4:50m, using 335 % CPU on average and a maximum of 1190 MB RAM.

Restored backup to a folder (lives on internal SSD):

Kopia: 21:35m, using 81 % CPU on average and a maximum of 1356 MB RAM.
Restic: 5:54m, using 80 % CPU on average and a maximum of 486 MB RAM.

Borg is slightly faster than restic when not compressing the backup, but it's noticeably slower than restic when compression is turned on. This is to be expected since borg is fundamentally single-threaded whereas restic and kopia are multi-threaded. Borg has an option to disable encryption whereas kopia and restic enforce encryption even if you don't need it (which wastes a lot of CPU time). Kopia is the fastest which is also expected since it backs up less metadata than restic and uses a cache that's ≈7x bigger than restic's cache.

Backup /foo/bar/data

Borg

Borg will only back up the folder "data" and all its contents, it will not back up "foo" or "bar".
However, borg preserves the file system structure /foo/bar/data by re-creating "foo" and "bar" if they don't already exist when restoring, but with the access permissions of the restore process. They'll be new folders, not the original "foo" and "bar" folders.

Restic

Restic will backup all 3 folders "foo", "bar" and "data".

Kopia

Kopia will only backup the contents within "data". It will not back up "foo", "bar" and "data" and will not preserve the file system structure /foo/bar/data.

Backup .

Borg

Borg will backup the current working directory with all its contents, but the contents and the current working directory (.) will be listed at the same folder-level (`borg list`). Additionally, the current working directory's name will not be preserved, it will be saved as "." When restoring the backup, the current working directory will be overwritten by the backed up working directory and all contents of the current working directory will be overwritten by the contents of the backed up working directory.

Restic

Using "." as a backup source is interpreted as all files within the current working directory, but not the current working directory itself.

Kopia

Using "." as a backup source is interpreted as all files within the current working directory, but not the current working directory itself.

Exclude/Include path pattern

Borg

Exclude/Include paths passed to Borg are always relative to the "active recursion root" (the path argument(s) at the end of `borg create`). A leading / will be stripped, making absolute file paths relative to the active recursion root. Exclude/Include paths pattern always match the whole path:
For example, "test" will not match all folders called "test". It will only match the top-level "test" folder.

Restic

When backing up, exclude/include paths with a leading / are relative to the root directory of the file system. When restoring, exclude/include paths with a leading / are relative to the root directory of the backup. Path patterns without a leading / are not matched against the whole path. For example, "test" matches all files called "test" instead of matching a file "test" relative to some other folder.

Kopia

Absolute exclude/include paths are always relative to the backup's root directory. Restic ignores a trailing slash, but kopia will only match directories in that case.

Restore backup

Borg

borg extract does not synchronize the target directory with the backup folder, meaning extranous files in the target directory will not be deleted unless the backup folder contains matching file paths. This makes borg useless for backups of the whole system.

Restic

Restic cannot preserve a file's change and birth timestamps. It also doesn't preserve the access timestamp by default, meaning it only preserves the modification timestamp by default.

Kopia

Kopia does not support exclude paths when restoring a backup (unlike restic).
Kopia does not preserve SUID/SGID and other rather linux-specific metadata (unlike restic).
Kopia does not preserve hard links (unlike restic).
Kopia cannot create sparse files when restoring a backup (restic can, but it's a rather bad implementation which often creates sparse files bigger than the original).

Conclusion

Use restic for backing up and restoring a linux system, borg and kopia can (and probably will) corrupt the system:
Borg cannot delete extranous files when restoring. Also, it's the slowest of the 3 because it's single-threaded.
Kopia is fast, but it does not support advanced file system features that are essential to a linux system.
It may not even be that fast if it would collect all the the additional metadata that restic is collecting.
Since exclude paths aren't supported when restoring, you're also not able to restore the system while it is running (because it would try to empty /run, /dev, ...) so you would need to boot into another system for a restore which is time consuming and annoying.

I would like to find a program that creates backups of entire volumes. In my case that means the C-drive of a Windows 11 system. I have no interest in file backup capabilities as I already have a system to do that. But I want something that will take the entire OS and save it in a restorable format, probably as a file on another local device I have or on a flash drive.

Cheap is nice, but I am willing to pay for something reliable. But not a monthly fee, as I likely will only use it a couple of times.

I used to use Acronis for this because it came with my WD drive and worked well. But now they want a monthly fee and I am not prepared to buy anything on a monthly basis.

Any thoughts on what I should use?

Built a little tool that keeps your own local copy of your Telegram chats, so nothing ever really disappears. Channels can go down, messages can get deleted (even by the other person), and you'll still have everything saved on your side. Early and open source; curious what people think.