OMG - I can't believe I did this

[Roken]

I have literally trashed all my media and my PI NFS boots.

So long version, I was upgrading to a 1TB drive from a 500Gb, since space was getting low. rsync is the better option, but I decided to clean the drive first. I cleaned twice, and have cleared both copies. All videos (replaceable), music (much has survived on another machine), and photos, some of which are irreplaceable.

Don't do system maintenance before noon on Sundays

[joan]

Oh dear. I hope you didn't lose anything too precious.

[Roken]

I did, unfortunately. However, some of the most precious may be on my phone, or other people's phone. I've lost 40 years of memories, but I may be able to get back the most important by contacting the right people, checking my social media posts, going back to old digital cameras and see what survived, etc.

There are two photos from more than 40 years ago of me and my better half, that I really hope she still has.

[Heater]

Oh boy. Sorry to hear that.

You have made me very nervous.

You see, I have for a long time relied on all my files going back to 1995 or so being kept on two or more machines at different physical locations. Home, office, elsewhere...

Well, what with the recent closure of our office and other circumstance I now find myself looking at a a pile of hard drives, from those machines, all sitting on the same shelf in front of me!

You have prompted me to do something about it before something bad happens. NOW!

Hope you find your precious files some place.

[epoch1970]

If you stop using the drives right now, perhaps there is still some hope, normally deleting will not overwrite files, just mark the corresponding blocks free for use.
There is at least one undelete utility for ext3 (4?) but I don’t think I ever had luck with it.
I had luck with photorec, which reads blocks, tries to identify files formats and copies them elsewhere. In the process you lose file name, location, but it is possible to recover files, as long as they have not been overwritten in the meantime...

[thatchunkylad198966]

Oh; I remember this too well!
I had years and years of data stored on a PC, I messed up and formatted the wrong drive whilst trying to install Linux. (this was way back!) and I never did managed to recover the data, luckily, Google photo's auto backup if you've got it set which I have so had my pictures saved but stuff like ripped dvd's, .iso files and everything was gone.

To this day, I'm still finding things that I need that I lost.

I feel for you, man.

[HawaiianPi]

I know that feeling. Did something similar myself many years back. I forget the exact details (dissociative amnesia?), but I wiped nearly 4TB of data with an errant command, and spent days trying to recover it from the drives to no avail. Don't beat yourself up, it happens to the best of us, even to someone like me who shouts from his soapbox about the virtues of multiple backups.

So what about those multiple backups ...

When I went to recover my deleted data from a redundant backup on an external HDD, the drive decided to crap out.

So the combination of an ID-10-T error and Murphy's law did me in. Fortunately I had some of it archived on many smaller drives, some on cloud services, and some shared with friends. After a lot of work collecting and comparing versions I was able to recover a good percentage of it. So hopefully you will be able to do the same.

Good luck.

[DougieLawson]

I have literally trashed all my media and my PI NFS boots.

I did it to a mainframe filesystem once. They gave me UID=0 and I managed to do a rm -rf * in the root.

The last failure (lost lots of stuff including photos) was done by my eldest son. He had some electricians test equipment and decided to try it on our house. On the fourth test in a row the hard drive in my NAS expired. It had three partitions on it. The latest backup (just after I'd cleaned up the backup drive) had failed after partition #1 (and I'd not noticed). Lost two thirds of the contents that way.

[Heater]

Over the years I have come to think that backups are a really bad idea.

Here are the problems I have with backups:

1) As soon as you put your data on a backup tape, disk, whatever and stash that away in a safe place it's as good as dead. You forget about it till you need it. Media rots while you are not paying attention. Grease in drives solidifies. So often we hear of failure to recover backups.

2) Backing up too much stuff. I'm looking at that collection of backup hard drives I mentioned earlier. Grief there are billions of files in there. I'm very sure I don't need or want most of it. Important stuff is as good as lost just now. It takes for ever to find it!

I start to think it would be better not to have backups. Rather one would keep data live at all times. It would live on machines the are geographically separated. Those machines would be permanently on. Slowly but surely they would talk to each other and verify that everyone has good copies of all the files. Any failures would be notified immediately.

Nothing would ever get deleted. Copies are read only. Space would be a problem. Thus making one think before simply backing up the entire content of a drive from a dead PC or whatever, just in case.

[Imperf3kt]

I've never accidentally formatted a hard drive, but I did once destroy one

I decided it was time to backup my 1.5TB WD green because it was a pile of crap.
I went out and bought a reliable 2TB Seagate Barracuda.

After setting everything up and beginning the transfer, I say impatiently and watched the data agonisingly slowly trickle onto the new HDD.. Something wasn't right with the USB on my desktop PC, so I went to get my laptop instead.

Well that did it... I promptly tripped over the power cable sending the hard drive flying across the room mid-read.
The drive still works, but I cannot get any computer to mount the partitions.

I lost 98% of the contents, but the problem was that I don't know what those contents were.

[LTolledo]

The closest thing that happened to me was when my 2TB RAID1 NAS (2x2TB in mirrored configuration) suddenly became unavailable.
checking the error codes, it seemed that one of the hard disk was has failed.
set me to panic mode how to recover my data?
removed the HDDs from the RAID enclosure and plugged known good HDD of the pair to a PC as external USB drive to see if I can move it to a 3TB HDD I have on spare.
cant be read, as the file system was not readable under windows.
panic attack2. how to recover my data?
searched for and tested data recovery software. eventually got one and paid license.
took 3 days to scan the files on the known good HDD and another 2 days to recover the data to the 3TB HDD (got 99% of all contents)

I've since upgraded the RAID1 NAS to 4TB (2x4TB) and manually mirrored it to a separate USB HDD using SyncToy.

nowadays am using another RPi2B with 4TB HDD, running timed mirroring of the RAID1 NAS, so I dont have to do it manually via synctoy on a win10 laptop.

[rpdom]

I backup my systems weekly every Sunday morning.

On the first Sunday of each month I take a full backup of each system to a server with sufficient disk space.

On the remaining Sundays I use an incremental backup to only store the files that have changed since the previous backup.

I also make additional backups of my databases.

The backups are then copied with rsync to external drives on two other systems.

Rotate those files every other month.

I'm covered. I've had to rely on backups only a few times in the last 20 years or so, but am glad I had them. They've saved me a lot of work over a few hard disk failures and the occasional oops!

Back in my Mainframe days we used a more complex system of backups and verification.
1. Backup the the main system disks to two sets of tapes - Master and Copy.
2. Shut down the mainframe and remove the disks.
3. Get the alternate set of disks from the fire safe and mount those.
4. Boot the system up and restore from the new backup tapes.
5. Put the first set of disks in the fire safe.
6. Send the backup copy tapes off site to secure storage.

This meant we knew we had three good backups of the system - a set of working disks and two sets of tapes.
(Until one idiot forgot to swap some of the disks and somehow ended up with an unbootable system. IT took us a few hours, but we managed to recover from that one with no loss of data or user downtime)