You are not logged in.
This has happened to me 3 times. I use my system normally, but occasionally on a new day when I boot, I get the screen saying "Failed to start. Dependency failed /home" and when I run fsck on it and restart, I am stuck at grub rescue.
I don't know why this error keeps happening to me even though I don't do anything out of ordinary on my system.
It looks like my root partition has only one directory, which is lost+found.
I don't want to reinstall my system again. Can someone please help me and share the reason why this keeps happening?
Thank you!
Offline
Post your fstab and
lsblk -f
in code tags I'm assuming you aren't using stable identifiers in your fstab and thus land on something else, see: https://wiki.archlinux.org/title/Persis … ice_naming
Online
lsblk -f
sdb
├─sdb1
│ vfat FAT32 NO_LABEL 2550-6A12
├─sdb2
│ swap 1 41ae4bf0-eb8b-404f-a25f-c90466cc6fa1
├─sdb3
│ ext4 1.0 3f8fe24a-2f16-4cc7-b941-294c4a693273 81.5G 12% /mnt
└─sdb4
ext4 1.0 35a6b520-f7e4-4cdf-bc15-eee2dd767c1e
sda is a hard disk which I currently don't use, nor have I ever mounted it in my system.
Whats the command for fstab?
Edit: sdb1 is boot, 2 is swap, 3 is root, and 4 is home.
Also, I don't know anything about stable identifiers.
Last edited by tsratsra (2022-01-14 15:32:16)
Offline
How did you install Arch? Which instructions did you follow? Your fstab is a file under /etc/fstab.
Online
fstab
# <file system> <mount point> <type> <options> <dump> <pass>
UUID=2550-6A12 /boot/efi vfat umask=0077 0 2
UUID=41ae4bf0-eb8b-404f-a25f-c90466cc6fa1 swap swap defaults,noatime 0 0
UUID=3f8fe24a-2f16-4cc7-b941-294c4a693273 / ext4 defaults,noatime 0 1
UUID=35a6b520-f7e4-4cdf-bc15-eee2dd767c1e /home ext4 defaults,noatime 0 2
tmpfs /tmp tmpfs defaults,noatime,mode=1777 0 0
Edit: this output is from command "cat /mnt/lost+found/#524289/fstab" as I don't have any file in root partition other than lost+found.
Last edited by tsratsra (2022-01-14 16:03:05)
Offline
You should fix the order here so that / comes first, in any case, maybe post the output of
smartctl -a /dev/sdb #Or whatever the relevant device ends up being
(additional to make sure the results are current you might want to run a SMART check from the ISO and after the mentioned testing time has elapsed post the output (... again, but for immediate diagnosis the current output might help already: https://wiki.archlinux.org/title/List_o … n_services
Last edited by V1del (2022-01-14 16:14:44)
Online
Offline
That... doesn't look good. You probably don't want to rely on that drive anymore. Don't try to write to it, to try and recover data get yourself an image onto a known working drive with something like dd/ddrescue and operate on that to get to data. You definitely shouldn't try to use it for anything important as it stands.
Online
What!!! I bought this ssd like 2 years ago!!
Edit: can you please tell me how to recover it??
Last edited by tsratsra (2022-01-14 16:58:30)
Offline
Maybe wait for some more opinions but there are multiple red flags here, from the reallocated sector counts, to the read errors it actively registered, a spare block count of 3 all things that don't read healthy at all.
https://wiki.archlinux.org/title/File_recovery and https://wiki.archlinux.org/title/Disk_c … g_ddrescue contain some relevant pointers to help in recovery.
Last edited by V1del (2022-01-14 17:03:01)
Online
Okay Thanks a lot for your support!! I want to recover sdb3 to sdb3, can I do that? I have spent a lot of time configuring things and I don't want to do stuff again if possible. Can I try to get things normal as before?
Offline
Error 1618 occurred at disk power-on lifetime: 1578 hours (65 days + 18 hours)
9 Power_On_Hours 0x0032 100 100 000 Old_age Always - 4826
I'm not saying that the drive is ok (the smart data does't look healthy) but the read errors were quite a while ago and none happened recently.
There's however also
Warning: ATA error count 1618 inconsistent with error log pointer 5
Either the controller is fried or the smart data of the drive is generally not very reliable.
this output is from command "cat /mnt/lost+found/#524289/fstab" as I don't have any file in root partition other than lost+found.
That's certainly extremely bad. Like, really badly bad.
Do you do any of https://wiki.archlinux.org/title/Solid_state_drive#TRIM ?
You should rescue all the data you can to a trusted drive and then submit this one to https://wiki.archlinux.org/title/Badblocks to see how bad things really are.
Edit:
I want to recover sdb3 to sdb3
Err… what?
Edit #2:
https://wiki.archlinux.org/title/File_recovery but for the moment you should forget the idea to write anything on that drive.
Last edited by seth (2022-01-14 17:11:54)
Offline
Again, active writes to that same drive can make stuff worse potentially leading you to be unable to access the data at all. You need to get a distinct hard drive/ssd and write the data you want there. If you do not currently have a backup, now you know why people have backups.
Online
Err… what?
I am sorry. What I meant is I want to try to recover the current partition ie root sdb3 , to the same partition ie root sdb3, to try to boot and see if all files are back to normal.
Offline
Again, active writes to that same drive can make stuff worse potentially leading you to be unable to access the data at all. You need to get a distinct hard drive/ssd and write the data you want there. If you do not currently have a backup, now you know why people have backups.
I am a student, and my financial situation is not good. All my very important data is on cloud. I want to use the current ssd as long as possible, Its okay if I have to recover it and let it die again and again. Can you help me recover and get back to normal.
Offline
That's not a thing and as V1del and myself have pointed out: forget about writing on that disk for now.
Rescue the *valuable* data. That's not your OS, that's not your… well, maybe also your porn collection, but mostly family pictures, texts/thesīs you wrote for school/university etcetc.
Then test the quality of the drive and only if you find that it's still reliable and the data loss is a user error (eg. bogus trimming or you rm'd too much) think about installing an OS again. And I mean "install" because that's gonna turn out easier that sticking together some file fragments and figuring which ones are still intact.
If your actual root partition only contains a lost+found directory, something really bad will have happened.
Offline
That's not a thing and as V1del and myself have pointed out: forget about writing on that disk for now.
Rescue the *valuable* data. That's not your OS, that's not your… well, maybe also your porn collection, but mostly family pictures, texts/thesīs you wrote for school/university etcetc.
Then test the quality of the drive and only if you find that it's still reliable and the data loss is a user error (eg. bogus trimming or you rm'd too much) think about installing an OS again. And I mean "install" because that's gonna turn out easier that sticking together some file fragments and figuring which ones are still intact.
If your actual root partition only contains a lost+found directory, something really bad will have happened.
Okay, thanks for your support. This has happened multiple times, and this time I had trim turned on weekly on both / and /home.
Last edited by tsratsra (2022-01-14 17:22:47)
Offline