[Bug 804072] New: GRUB2 + kernel 3.4.28-2.20-desktop fails to boot with Bad Magic for RAID
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c0 Summary: GRUB2 + kernel 3.4.28-2.20-desktop fails to boot with Bad Magic for RAID Classification: openSUSE Product: openSUSE 12.2 Version: Final Platform: x86-64 OS/Version: openSUSE 12.2 Status: NEW Severity: Critical Priority: P5 - None Component: Bootloader AssignedTo: jsrain@suse.com ReportedBy: peter@peterchambers.me.uk QAContact: jsrain@suse.com Found By: --- Blocker: --- Created an attachment (id=524920) --> (http://bugzilla.novell.com/attachment.cgi?id=524920) /boot/grub2/grub.cfg User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:14.0) Gecko/20100101 Firefox/14.0.1 Boot error (stop) using Grub2, kernel vmlinux-3.4.28-2.20-desktop.gz with system laid out as below. At start of problems: #fdisk -l /dev/sda Device Boot Start End Blocks Id System /dev/sda1 * 2048 145407 232448 83 Linux /dev/sda2 145408 4354047 1928192 82 Linux swap / Solaris /dev/sda3 4354048 976773119 486209536 fd Linux raid autodetect # fdisk -l /dev/sdb Device Boot Start End Blocks Id System /dev/sdb1 2048 145407 232448 83 Linux /dev/sdb2 145408 4354047 1928192 83 Linux /dev/sdb3 4354048 976773119 486209536 fd Linux raid autodetect This is /dev/sda1 /boot 70 Mb ext4 /dev/sda2 swap 2 Gb /dev/sda3 md1 RAID element supporting LVM 'system2' /dev/sdb1 ext4 70 Mb (match /boot size) /dev/sdb2 empty 2 Gb (match swap) /dev/sdb3 md1 RAID element supporting LVM 'system2' LVM system2 group has two ext4 file systems: 'root' and 'home' Originally set up with Yast2 installer from 12.2 DVD downloaded as ISO image from opensuse. Reproducible: Always Steps to Reproduce: Steps to reproduce: 1. start with working system using vmlinux-3.4.6-2.10-desktop.gz 2. update to kernel vmlinux-3.4.28-2.20-desktop.gz - as routine Online Update. 3. reboot system on request 4. system re-started and booting kernel reports "bad magic" for /dev/md1 (mentions 0.9 metadata) Reproducible: twice (once on original system, again when 12.2 installed from DVD and Online Update used) Actual Results: System hangs with kernel trace on video screen reporting bad magic with /dev/md1. Expected Results: Should have booted into updated system. RAID information # mdadm --detail /dev/md1 /dev/md1: Version : 1.0 Creation Time : Sat Dec 1 21:01:01 2012 Raid Level : raid1 Array Size : 486209344 (463.69 GiB 497.88 GB) Used Dev Size : 486209344 (463.69 GiB 497.88 GB) Raid Devices : 2 Total Devices : 2 Persistence : Superblock is persistent Intent Bitmap : Internal Update Time : Thu Feb 14 21:17:55 2013 State : active Active Devices : 2 Working Devices : 2 Failed Devices : 0 Spare Devices : 0 Name : linux:1 UUID : b4e13cf0:82bf8c6c:e3b1e879:77979528 Events : 538324 Number Major Minor RaidDevice State 0 8 3 0 active sync /dev/sda3 2 8 19 1 active sync /dev/sdb3 Hypotheses investigated after bugzilla and Google searches a. RAID 0.90 metadata might be an issue - established 1.0 metadata is on /dev/md1. b. space for /boot may be too small during bootloader making by Yast2 (e.g. initrd) so traded off swap space for bigger (242 Mb /boot volume), see below for new layout, re-made swap and ext4 partitions and restored data from backups. c. possibility that "core.img" may need space below /dev/sda1, see below for new layout approximately 16 Mb provided between MBR and /dev/sa?1. d. missing mdraid modules (discovered mdraid1x.mod in /boot/grub2/i386-pc/) e. initrd missing something (not investigated in depth) f. MBR/bootloader needed fixing (on /dev/sda), tried recipe as follows: power on machine, boot into 12.2 DVD rescue system vgchange -ay system2 # activated system2 and proved RAID works from DVD rescue system mount /dev/mapper/system2-root /mnt # works, can see '/' inside LVM mount /dev/sda1 /mnt/boot # works, can access files in '/boot' mount --bind /dev /mnt/dev chroot /mnt # gives me chroot environment CLI grub2-install /dev/sda # refreshes MBR, does not fix problem References see: http://forums.opensuse.org/english/get-technical-help-here/install-boot-logi... https://forums.opensuse.org/english/get-technical-help-here/install-boot-log... All current bugzilla issues with "raid" in subject line, 785226 seems possibly related A Google search turns up various similar unfixed issues with Gentoo and other distros, going back into 2012, no fixes. === the new HDD layout === #fdisk -l /dev/sda Disk /dev/sda: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors Units = sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 4096 bytes I/O size (minimum/optimal): 4096 bytes / 4096 bytes Disk identifier: 0x000a94df Device Boot Start End Blocks Id System /dev/sda1 * 32768 497663 232448 83 Linux /dev/sda2 497664 4354047 1928192 82 Linux swap / Solaris /dev/sda3 4354048 976773119 486209536 fd Linux raid autodetect # fdisk -l /dev/sdb Disk /dev/sdb: 500.1 GB, 500107862016 bytes 255 heads, 63 sectors/track, 60801 cylinders, total 976773168 sectors Units = sectors of 1 * 512 = 512 bytes Sector size (logical/physical): 512 bytes / 4096 bytes I/O size (minimum/optimal): 4096 bytes / 4096 bytes Disk identifier: 0x00049e8d Device Boot Start End Blocks Id System /dev/sdb1 32768 497663 232448 83 Linux /dev/sdb2 497664 4354047 1928192 83 Linux /dev/sdb3 4354048 976773119 486209536 fd Linux raid autodetect == relevant lines from device.map (hd0) /dev/disk/by-id/ata-WDC_WD5003AZEX-00K1GA0_WD-WMC1S0816380 (hd1) /dev/disk/by-id/ata-WDC_WD5003AZEX-00K1GA0_WD-WMC1S0749794 == key lines from grub.cnf (full file available later) echo 'Loading Linux 3.4.28-2.20-desktop ...' linux /vmlinuz-3.4.28-2.20-desktop root=/dev/mapper/system2-root video=1920x1080 resume=/dev/disk/by-id/ata-WDC_WD5003AZEX-00K1GA0_WD-WMC1S0816380-part2 splash=silent quiet showopts echo 'Loading initial ramdisk ...' initrd /initrd-3.4.28-2.20-desktop -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c Lukas Ocilka <locilka@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |snwint@suse.com AssignedTo|jsrain@suse.com |mchang@suse.com -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c1 --- Comment #1 from Michael Chang <mchang@suse.com> 2013-02-22 09:46:13 UTC --- Thanks to rich collection of the info here. It's good for us all to track the mdraid metadata 0.9 related issues here. To be honest this beyond my ability to fix. :( -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c2 --- Comment #2 from Michael Chang <mchang@suse.com> 2013-02-22 10:00:19 UTC --- Similar to bnc#785226. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c3 Andrey Borzenkov <arvidjaar@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEW |NEEDINFO InfoProvider| |peter@peterchambers.me.uk --- Comment #3 from Andrey Borzenkov <arvidjaar@gmail.com> 2013-02-22 10:07:18 UTC --- (In reply to comment #0)
4. system re-started and booting kernel reports "bad magic" for /dev/md1 (mentions 0.9 metadata)
If kernel boots, bootloader task is over. Why do you think it is grub2 related? Attaching screenshot at the point kernel stops would be useful. Also try to boot using shell=1; it should drop you into shell in initrd, where you can try to collect more information at the time of failure. What is the hostname of your system? I think I know what happens. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
https://bugzilla.novell.com/show_bug.cgi?id=804072 https://bugzilla.novell.com/show_bug.cgi?id=804072#c4 Jeffrey Cheung <jcheung@suse.com> changed: What |Removed |Added ---------------------------------------------------------------------------- Status|NEEDINFO |CLOSED CC| |jcheung@suse.com InfoProvider|peter@peterchambers.me.uk | Resolution| |WONTFIX --- Comment #4 from Jeffrey Cheung <jcheung@suse.com> 2014-02-11 04:29:10 UTC --- With the release of the gnumeric on January 27th, 2014 the SUSE sponsored maintenance of openSUSE 12.2 has ended. openSUSE 12.2 is now officially discontinued and out of support by SUSE. So, I am going to close this issue as WONTFIX. Please to create a new one if openSUSE 13.1 encounter the same problem. -- Configure bugmail: https://bugzilla.novell.com/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@novell.com