[Bug 10916] New: apparent interaction bug between ACPI and RAID rebuild in Linux

[bugzilla-daemon at bugzilla.ubuntu.com]

Please do not reply to this email.  You can add comments at
http://bugzilla.ubuntu.com/show_bug.cgi?id=10916
Ubuntu | linux

           Summary: apparent interaction bug between ACPI and RAID rebuild
                    in Linux
           Product: Ubuntu
           Version: unspecified
          Platform: i386
        OS/Version: Linux
            Status: UNCONFIRMED
          Severity: major
          Priority: P2
         Component: linux
        AssignedTo: fabbione at ubuntu.com
        ReportedBy: bkuhn at ebb.org
         QAContact: kernel-bugs at lists.ubuntu.com


Installing in server mode from Hoary ISO, I made the following configuration:
    * made /boot a RAID-1 ext3 filesystem across four partitions on four IDE disks
    * made / a RAID-5 ext3 filesystem across four other partitions on the same
four IDE disks

I proceeded with the install, and things seem nominal.  Then, when the install
reboots to complete, I get the following message when rebooting (turned off
quiet in grub to get more verbosity):

md: syncing RAID array md1
md: minimum _guaranteed_ reconstruction speed: 1000 KB/sec/dis
md: using maximum available idle IO bandwith (but not more than 2000000 KB/sec)
for reconstruction
md: using 128k window over a total of 151115520 blocks.
Stopping tasks: === [ there is a short pause ]
   stopping tasks failed (1 tasks remaining)

then, the system hangs.


I try the install again, but this time, I go into the console during the
install, and make sure the reboot doesn't happen until /proc/mdstat tells me the
rebuild of the new RAID-5 array is complete.  Then, the install proceeds
normally, and the system always boots normally, UNLESS it needs to do a rebuild
at boot time (because I force-failed a disk with mdadm, for example), and then I
get the same output as above.


However, if I set "acpi=off" in the boot string, the system boots normally and
begins the reconstruction as it normally should.

My theory is that this is an interaction bug between ACPI and the auto-start of
the RAID sync started by Linux.  This all seems to happen before init, so I
presume it's a Linux problem.

I've can get you lspci -v output from the machine in question, in case it is a
further interaction bug between specific hardware on the machine.  I'm happy to
help with more answers and experimentation.  For the moment, I'm just running
with ACPI as a work-around.

-- 
Configure bugmail: http://bugzilla.ubuntu.com/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the QA contact for the bug, or are watching the QA contact.