[Bug 1547193] [NEW] Impossible to answer "disk drive not ready" question on serial console

[Tore Anderson]

Public bug reported:

When booting a server that is suffering from a (probably unrelated) bug
in multipath-tools, I get the following questions posed on the serial
console during bootup:

The disk drive for /opt/vnx is not ready yet or not present.
Continue to wait, or Press S to skip mounting or M for manual recovery

Problem is, pressing "S" or "M" (even if followed by a newline) does
absolutely nothing on the serial console. It does work on the KVM
console, but that might not be available. You might very well need to
send someone to a physical data centre somewhere in the middle of
nowhere. (I ended up with powercycling the server, temporarily network
booting it into a ramdisk, mounting the rootfs and commenting out the
problematic entry from /etc/fstab, and rebooting again - which went
fine, allowing me to later mount the missing filesystem manually via
ssh.)

I found no way to make the boot finish or break out to a debug shell,
and it happens before sshd starts, so you're basically stuck with no
apparent way to recover.  For this reason I feel the bug is quite
severe.

For what it's worth I'm running 14.04.4 LTS, mountall package version is
2.53. The kernel command line is
«BOOT_IMAGE=/@/boot/vmlinuz-3.13.0-77-generic
root=UUID=2a141a43-97b0-4084-8816-ed1c41ac43e0 ro rootflags=subvol=@
console=tty1 console=ttyS0,115200n8».

Tore

** Affects: mountall (Ubuntu)
     Importance: Undecided
         Status: New

-- 
You received this bug notification because you are a member of Ubuntu
Foundations Bugs, which is subscribed to mountall in Ubuntu.
https://bugs.launchpad.net/bugs/1547193

Title:
  Impossible to answer "disk drive not ready" question on serial console

Status in mountall package in Ubuntu:
  New

Bug description:
  When booting a server that is suffering from a (probably unrelated)
  bug in multipath-tools, I get the following questions posed on the
  serial console during bootup:

  The disk drive for /opt/vnx is not ready yet or not present.
  Continue to wait, or Press S to skip mounting or M for manual recovery

  Problem is, pressing "S" or "M" (even if followed by a newline) does
  absolutely nothing on the serial console. It does work on the KVM
  console, but that might not be available. You might very well need to
  send someone to a physical data centre somewhere in the middle of
  nowhere. (I ended up with powercycling the server, temporarily network
  booting it into a ramdisk, mounting the rootfs and commenting out the
  problematic entry from /etc/fstab, and rebooting again - which went
  fine, allowing me to later mount the missing filesystem manually via
  ssh.)

  I found no way to make the boot finish or break out to a debug shell,
  and it happens before sshd starts, so you're basically stuck with no
  apparent way to recover.  For this reason I feel the bug is quite
  severe.

  For what it's worth I'm running 14.04.4 LTS, mountall package version
  is 2.53. The kernel command line is
  «BOOT_IMAGE=/@/boot/vmlinuz-3.13.0-77-generic
  root=UUID=2a141a43-97b0-4084-8816-ed1c41ac43e0 ro rootflags=subvol=@
  console=tty1 console=ttyS0,115200n8».

  Tore

To manage notifications about this bug go to:
https://bugs.launchpad.net/ubuntu/+source/mountall/+bug/1547193/+subscriptions