SLE 15 cannot boot production down

Hello,
I have a virtual machine in production that can no longer boot, I tried several tutorials on the internet but nothing

the host is ESXi 6.5
Suse enterprise linux 15 virtual machine
I tried to reinstall the grub with this support link: https://www.suse.com/support/kb/doc/?id=000016528, in Rescue mode.
already in step "cat / mnt / etc / fstab" the system crashes
then if I skip this step and launch the command "grub-install --root-directory = / mnt / dev / sda" I get this error "-bash grub-instal: command not found"

I also tried this tutorial: https://www.suse.com/support/kb/doc/?id=000018770
but the system crashes after this command "# chroot / mnt" with a dash that flashes at the bottom

please help me it is my production which is stopped


«1

Comments

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi and welcome to the Forum :)
    Seems to me grub is working since you were able to boot in rescue mode? Or did you boot from the install media?
    Sounds like you may have a full disk with no tweaking of snapper and btrfs?
    Can you post the output from;

    snapper list
    btrfs filesystem show /
    
  • edited July 1



    hello,
    thank you for reply
    see the output
    command not recognized

    Seems to me grub is working since you were able to boot in rescue mode? YES

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    OK, can you boot to grub and select the advanced options and boot from a previous snapshot? It will be read-only, but that's ok for the moment, can you confirm first?



  • We dont have snapshot
    when i boot from install disk see scrennshoot

  • malcolmlewismalcolmlewis Knowledge Partner
    edited July 2

    Hi
    You are booting from the installer, select the first option Boot from Hard Disk it will load grub again (the one on the HDD)?

  • yes i booted from hard disk but i have this screnn more than 20 minutes

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    That is plymouth, press the esc key when you see the init loading text. At grub you can temporarily disable by pressing a key to stop the countdown. Then press e to edit, at the line with linux (or linuxefi) add the following after the quiet;

    plymouth.enable=0
    
  • Hello,
    i dont know where i'll put this command

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    It's further down in the second screen shot you posted.

  • Hello,
    i added the i clic on F10
    the reboot by is same issue
    see below

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    So some AppArmor adjustments had been performed? It just sits there running? Can you try also adding apparmor=0 after the plymouth.enable=0 via editing grub at boot.

  • it seems that the last order that I have was not recorded, however I made "F10"

    ???

  • I also entered the grub, I launched the two commands
    but I still have the message

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    Yes, at each boot you will need to add the string as it's not saved (because can't boot). So in the grub menu you get too, in the advanced options there is no chance to select a previous snapshot? In the options can you select failsafe to boot from?

  • Hello,
    even with snapshots, I have a timer of several minutes




  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    When you say you have a timer of several minutes, does it come up now?
    If not, can you go back to say the 2020-06-11 pre one and see how that one goes?

  • hello,
    When you say you have a timer of several minutes, does it come up now? NO
    with the 2020-06-11 the boot process is blocked on this step below more than 4 hours

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    Are there earlier snapshots to try booting from?

  • Hi,
    Now iam trying to export the vmdk file
    there is a software to help me to get the data from the vmdk

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    I would suggest contacting SUSE Support and raising a Service Request before going down that road.

  • i dont have support suscription,
    Are there earlier snapshots to try booting from? NO, even if that worked, what about my database file
    is there another solutions we can try

  • malcolmlewismalcolmlewis Knowledge Partner
    edited July 3

    Hi
    Your database file should be on a separate partition on the system and type xfs, not btrfs?

    If so, should be able to boot into rescue mode and mount the one partition manually.

  • Hi,
    so if i can recover my database in rescue mode, my problem is already solved

    in which of these partitions can I find my database, so that I can mount it?

    how to manually recover this data, via ssh ??

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    Is it on /dev/dm-1 perhaps? If it's on /dev/sda1 that would appear to be btrfs (which is not designed for the likes of a database), then you will probably have issues...

    mount /dev/sda1 /mnt
    ls /mnt
    umount /mnt
    sync
    mount /dev/dm-1 /mnt
    ls /mnt
    
  • Hello,

    I copied the vmdk files with Winscp then with power iso I tried to read. but i have a message that says the files are corrupt.

    I'm out of ideas

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    What about my previous post with a manual mount?

  • Hi
    see below

  • malcolmlewismalcolmlewis Knowledge Partner

    Hi
    What about /dev/dm-1?

  • Hi
    see below

Sign In or Register to comment.