Dead system after update

[Grimble]

Downloaded the latest updates including kernel 4.4.55, but on restarting
one of my machines gave the "error 18 selected cylinder exceeds maximum
supported..." error. Googling around suggests it might be a hardware
problem, but Parted Magic can read /dev/sda, and is just carrying out an
extended test. Used Parted Magic to delete some of the earlier kernel
versions from /boot, but reboot gave a blank screen with a single
unmovable cursor.
Any ideas to correct the problem are welcome.

[Jim Beard]

Post by Grimble
Downloaded the latest updates including kernel 4.4.55, but on restarting
one of my machines gave the "error 18 selected cylinder exceeds maximum
supported..." error. Googling around suggests it might be a hardware
problem, but Parted Magic can read /dev/sda, and is just carrying out an
extended test. Used Parted Magic to delete some of the earlier kernel
versions from /boot, but reboot gave a blank screen with a single
unmovable cursor.
Any ideas to correct the problem are welcome.

Boot with a live CD or an emergency recovery disk and run fsck on your
boot drive. You might try fsck -y and see if it will automatically fix
something rather than just list errors.

One of several machines bad, the cylinder out of bounds, and a blank
screen all suggest hardware as the culprit.

If not a hardware problem, it might help if you identified whether you
boot to mbr or to UEFI, whether your file system is gpt, ext4, or
something else, and the video card and driver you use.

Cheers!

jim b.

[grimble]

Post by Jim Beard

Post by Grimble
Downloaded the latest updates including kernel 4.4.55, but on restarting
one of my machines gave the "error 18 selected cylinder exceeds maximum
supported..." error. Googling around suggests it might be a hardware
problem, but Parted Magic can read /dev/sda, and is just carrying out an
extended test. Used Parted Magic to delete some of the earlier kernel
versions from /boot, but reboot gave a blank screen with a single
unmovable cursor.
Any ideas to correct the problem are welcome.

Boot with a live CD or an emergency recovery disk and run fsck on your
boot drive. You might try fsck -y and see if it will automatically fix
something rather than just list errors.
One of several machines bad, the cylinder out of bounds, and a blank
screen all suggest hardware as the culprit.
If not a hardware problem, it might help if you identified whether you
boot to mbr or to UEFI, whether your file system is gpt, ext4, or
something else, and the video card and driver you use.
Cheers!
jim b.

Jim,
Thanks for the input. fsck -y did report (unspecified) errors on
/dev/sda8, but completed its 5 stages, and a subsequent run reported no
errors. Furthermore, an extended SMART run reported no errors. However,
looking through syslog during the last minutes of good operation,
actually while rpm was running, there were some suspicious messages that
I will need some time to understand. More tomorrow.
(FYI: mbr, ext4, nVidea)