System won't start after updates

Hello,

after installing updates today (sudo zypper up, I think lots of libs, grub etc., zypper told me to restart afterwards), my linux notebook can’t boot. I searched in the forums, but did not find anything that seems to fit.

Symptoms:
Booting takes long (spinning thingy), then breaks to ermergency console (“enter root pw or ctrl-d to continue”).

journalctl says

kernel: i915 0000:00:02.0: fb0: i915drmfb frame buffer device
systemd[1]: dev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device: Job dev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device/start timed out
systemd[1]: Timed out wating for device dev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device
systemd[1]: Dependency failed for Resume from hibernation using device /dev/disk/by-uuid/<uuid>
systemd[1]: systemd-hibernate-resume@dev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device failed with result ‘dependency’
systemd[1]: ev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device: Job dev-disk-by\x2duuid-…\x2ff…\x2da…\x2dd…device/start failed with result ‘timeout’
dracut-initqueue[303]: Warning: dracut-initqueue timeout - starting timeout scripts
<last msg repeats often>
dracut-initqueue[303]: Warning: Could not boot.

Resume from hibernation sounds weird, I did a reboot after installing updates.

Does anybody have an idea about what’s wrong with my box? Any suggestions would be appreciated!

Maybe your resume= is wrong in your grub-cmdline?

Post:

cat /proc/cmdline
cat /etc/fstab

Hi, thank you very much for the quick reply!


sh-4.4# cat /proc/cmdline
BOOT_IMAGE=/boot/vmlinuz-5.3.10-lp152.36-default root=UUID=<stuff> splash=silent resume=/dev/disk/by-uuid/<stuffY mitigations=auto quiet

sh-4.4# cat /etc/fstab
cat: /etc/fstab: No such file or directory


There’s only a fstab.empty file in /etc.

btrfs as filesystem?
can you boot into a snapshot?

btrfs as filesystem?
I think. I did not mess with installer default settings. Whole disk for opensuse. I think xfs is used somewhere as well? Not sure.

can you boot into a snapshot?
Unfortunately not. Same result. I tried restore mode as well, no result.

  1. At the Emergency console, enter the root password.
  2. At the CLI prompt, execute some Btrfs housekeeping:

 # btrfs balance start -musage=50 /
 # btrfs balance start -dusage=50 /
 # btrfs scrub start -B /

Once the housekeeping has completed, reboot:

 # systemctl reboot

Once the system has rebooted, please check that the following systemd services have been enabled:

  • btrfs-balance.timer
  • btrfs-scrub.timer

The associated systemd services, “btrfs-balance.service” and “btrfs-scrub.service” should have the status “static”.

  • If the system disk is an SSD then, also check that the “btrfs-trim.timer” service is enabled.

Once all that has been done, drop down to the systemd Rescue mode:

 # systemctl rescue

At the CLI prompt execute:

 # systemctl start btrfs-balance.service

Wait for the Btrfs Balance to complete.

If the system disk is an SSD then:

 # systemctl start btrfs-trim.service

Wait for the SSD Trim to complete.
Reboot.

Hi, wow, thank you.

Unfortunately,


# btrfs balance start -musage=50 /
# ERROR: not a btrfs filesystem: /

I wondered what the 15.2 Installer would say to current disk layout and booted from an installation medium (USB stick). I chose ‘Install’, and step “System probing” gave me an error message/window, saying


YAST2:
Command '/usr/bin/udevadm info '/dev/sda1'' failed:
stderr:
Unknow device, --name=, --path=, or absolute path in/dev/ or /sys expected.

exit code:
4

I checked if the ssd is still inserted correctly, looks good. HDD LED blinks a couple of times when trying to boot, but not very often.

Did sudo zypper up kill my ssd??


# /usr/bin/udevadm --query=property /dev/sda
<lots of stuff>
ID_MODEL = INTENSO_SSD_SATA_MODEL_III
<even more stuff>
#

→ Intenso SSD recognized.


# /usr/bin/udevadm --query=property /dev/sda1
Unknow device, --name=, --path=, or absolute path in/dev/ or /sys expected.

Did you check the results of “Repair Installed System”?

No, because I can’t find this.
Options when boot ing my Leap 15.2 USB stick:


Boot from Hard Disk
Installation
Upgrade
More ->
Rescue System
Boot Linux System
Check Installation Media
Memory TEst

There’s a “Rescue System” entry that asks for a rescue login. Accepts only root as user, no pw, says “have a lot of fun…” and drops me on a shell prompt.

Another thread here on the forum says there is no longer a rescue option, but selecting Upgrade will give the same result (with the need to install all updates again).

Where you can mount /dev/sda2 on /mnt.

  • Please note that, /dev/sda1 is usually the EFI partition used by UEFI/BIOS …
  • It may well be that /dev/sda4 or /dev/sda3 is the swap partition …
  • Your user XFS partition is possibly /dev/sda3 or /dev/sda4 …

but, before you mount the partitions on /dev/sda, you could execute “fsck” on each partition in turn – many fsck repairs can only be performed if the file system isn’t mounted:


 # fsck /dev/sda1
 # fsck /dev/sda2
 # fsck /dev/sda3
 # fsck /dev/sda4

@acheronHades:

If everything else fails, you may have to purchase a new SDD, perform a fresh installation on that drive and then, with the help of a USB case for your failing SSD, recover the user files from the SSD with the boot issue.


# fsck /dev/sda1
…
fsck.ext2: No such file or directory while trying to open /dev/sda1
Possibly non-existent device?
# lsblk
…
sda      8:0       0          223.6G 0          disk
sdb      8:16 …
    Sdb1…
    Sdb2…
...
# 

Did my box loose the partition table on sda?

Could gpart help in recreating it? gpart is available on the rescue system.


# fdisk -l /dev/sda
The primary GPT is corrupt, but the backup appears ok, so that will be used.
<stuff>
Device             Start    End      Sectors            Size      Type
/dev/sda1       2048    18431  16384              8M       BIOS boot
/dev/sda2       <stuff>                                   221.6G Linux filesystem
/dev/sda3       <stuff>                                   2G       Linux swap
# pvck /dev/sda2
Device /dev/sda2 not found
#

If fdisk can access stuff based on the backup GPT, do I need to tell the rest of the system to do the same?

The /dev/sda1 partition is a BIOS boot partition – the position is a little bit strange – it should start at sector 34 and, sector 2048 ain’t a multiple of 34 …

  • The parted
    flag bios_grub should be set (activated) on that partition.

Because /dev/sda1 is a BIOS Boot partition, it ain’t formatted – fsck has absolutely no support for BIOS Boot partitions and, doesn’t need any support for them …

If you don’t have any Logical Volumes, pvck isn’t of any use …

GRUB should have embedded a core.img file in the /dev/sda1 partition – you may have to reinstall GRUB:

 # grub2-install /dev/sda

Running fsck on /dev/sda2 should repair any issues you have on that partition.

Repairing the swap partition on /dev/sda3 is not self-explanatory – there ain’t no tools to repair a swap partition …


 # swapoff /dev/sda3
 # mke2fs -c /dev/sda3
 # mkswap -c /dev/sda3
 # swapon /dev/sda3

The two file system creation commands “mke2fs” and “mkswap” with the “-c” options should find any issues with the SSD – if both report errors, you’ll need to purchase a new SSD …


# fsck /dev/sda1
…
fsck.ext2: No such file or directory while trying to open /dev/sda1
Possibly non-existent device?
# grub2-install /dev/sda
<some error about not finding paths>

I solved it this way:
all commands fail with various errors regarding “no /dev/sda1 or /dev/sda2” etc. Only fdisk is smart enough to see the primary gpt is broken and uses the backup one.
I simply openend fdisk, it gave the warning again (“The primary GPT is corrupt, but the backup appears ok, so that will be used.”), and then I told it to w (write) the gpt.

System works again lol!

Thank you all for your ideas and suggestions, and MANY thanks for staying with me and my problem 15+ forum entries long!!

Sorry, typo in last post:


# fsck /dev/sda2
…
fsck.ext2: No such file or directory while trying to open /dev/sda1
Possibly non-existent device?
# grub2-install /dev/sda
<some error about not finding paths>