Post resizing the disk of a newly deployed SSB VM, the core partition is not expanded consistently. The VM becomes unusable after reboot. (4375318)

Post resizing the disk of a newly deployed SSB VM, the core partition is not expanded consistently. The VM becomes unusable after reboot.

When a user wants to use the disk resize feature on a freshly deployed SSB VM, the core partition will not be expanded, rendering the VM unusable after reboot.

Affected versions: SSB 7.0.1 LTS, 7.0.2 LTS, 7.3.0F
Instances upgraded from the original install of 7.0.0 LTS or 7.2.0F and earlier are NOT affected, only fresh installs.
Affected platforms: hypervisors where the installation is made via ISO file (VMWare ESXi, Hyper-V).
Azure and AWS deployments are NOT affected.

 

Product Defect ID 453276

Please choose your initial disk size carefully.
The issue is under investigation and will be fixed in upcoming releases.

Workaround. The symptom will be that the core firmware will not start on the SSB if the described defect is triggered by a disk resize. The following will be seen in the (virtual machine's) console:



The key lines are the following messages from the above screenshot:

e2fsck: No such file or directory while trying to open Can not determine device of core firmware read-write layer

Possibly non-existent device?

mount: /mnt/firmware: wrong fs type, bad option, bad superblock on none, missing codepage or helper program, or other error.

ERROR: Core firmware preparing failed.

When the user hits Enter in this console, they will be able to enter the username and password to log in. Upon login, they will see the following screen:

The user has to hit Enter once again to get into the fallback boot shell:

Execute the following command to see the partition table of the SSB's OS disk:

sgdisk -p /dev/sda



If there is no partition #4 with the name core, then it was deleted according to the bug description (like in the screenshot above).

To remediate the situation, execute the following chained-command:

/usr/sbin/sgdisk --new '4:12587008:' --typecode '4:8300' --change-name '4:core' /dev/sda && /usr/sbin/partprobe /dev/sda && /usr/bin/timeout 60s /usr/bin/bash -c 'until ! lslocks -nuo path | grep -qx /dev/sda ; do sleep 1 ; done' && /usr/sbin/resize2fs /dev/sda4 && /usr/sbin/e2fsck -f -y /dev/sda4 && /usr/sbin/reboot

The chained command will do the following:

1.    Re-create the previously deleted partition of the core firmware (starting from the correct sector number).
2.    Inform the OS of the partition table changes.
3.    Wait until locks regarding the re-created partition are released (if there is any).
4.    Resize the ext4 filesystem of the core firmware.
5.    Fix any filesystem errors of the core firmware (if there is any).
6.    Reboot the SSB (virtual machine).

453276