Bug#385951: Fail to scan array in some case when partition is at the end of the disk

Klaus Ethgen Klaus at Ethgen.de
Mon Sep 4 08:05:22 UTC 2006


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Package: mdadm
Version: 2.5.2-7
Severity: critical

Tha actual mdadm in testing makes the whole system to unbootable if on
raid slice is on the end of a disk. (In some cases)

My config:
- ---mdadm.conf---
DEVICE partitions
ARRAY /dev/md0 level=raid1 num-devices=2 UUID=88cf7fb7:6fab12d7:b713c983:af6eaca5
MAILADDR root
- ----------------

- ---hdc---
Disk /dev/hdc: 60.0 GB, 60060155904 bytes
16 heads, 63 sectors/track, 116374 cylinders
Units = cylinders of 1008 * 512 = 516096 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hdc1   *           1         496      249983+  83  Linux
/dev/hdc2             497        1488      499968   82  Linux swap / Solaris
/dev/hdc3            1489      116374    57902544   fd  Linux raid autodetect
- ---------
- ---hdd---
Disk /dev/hdd: 320.0 GB, 320072933376 bytes
255 heads, 63 sectors/track, 38913 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hdd1               1        9728    78140159+  8e  Linux LVM
/dev/hdd2            9729       19456    78140160   8e  Linux LVM
/dev/hdd3           31705       38913    57906292+  fd  Linux raid autodetect
- ---------

- ---/proc/mdstat---
Personalities : [raid1] 
read_ahead 1024 sectors
md0 : active raid1 hdd3[0]
      57902464 blocks [2/1] [U_]
      
unused devices: <none>
- ------------------

Please note that I had to run the raid in degraded mode as /dev/hdc3
cannot be in the array (See below).

When I do a "mdadm --assemble --scan --auto=yes" (as done in
/etc/init.d/mdadm-raid) then I get the following error:

- ---
mdadm: no recogniseable superblock on /dev/hda2
mdadm: /dev/hda2 has wrong uuid.
mdadm: no recogniseable superblock on /dev/hda1
mdadm: /dev/hda1 has wrong uuid.
mdadm: no recogniseable superblock on /dev/hda
mdadm: /dev/hda has wrong uuid.
mdadm: no RAID superblock on /dev/hdd2
mdadm: /dev/hdd2 has wrong uuid.
mdadm: no RAID superblock on /dev/hdd1
mdadm: /dev/hdd1 has wrong uuid.
mdadm: no RAID superblock on /dev/hdd
mdadm: /dev/hdd has wrong uuid.
mdadm: no RAID superblock on /dev/hdc2
mdadm: /dev/hdc2 has wrong uuid.
mdadm: no RAID superblock on /dev/hdc1
mdadm: /dev/hdc1 has wrong uuid.
mdadm: no RAID superblock on /dev/vg1/lv_hathi
mdadm: /dev/vg1/lv_hathi has wrong uuid.
mdadm: no RAID superblock on /dev/vg1/lv_misc
mdadm: /dev/vg1/lv_misc has wrong uuid.
mdadm: no RAID superblock on /dev/vg1/lv_mirror
mdadm: /dev/vg1/lv_mirror has wrong uuid.
mdadm: WARNING /dev/hdc3 and /dev/hdc appear to have very similar superblocks.
      If they are really different, please --zero the superblock on one
      If they are the same, please remove one from the list.
- ---

Cleaning the superblock of /dev/hdc also clean the one of /dev/hdc3 and
so leaving the raid in a degraded state. But at least I can now boot the
system after running the md0 by hand.

Note that the partition /dev/hdd3 is also at the end of the disk but do
not make problemes.

This is a very critical bug and should be fixed in etch (I think, this
is release critical!!!)

- -- Package-specific info:
- --- mount output
/dev/hda1 on / type ext3 (rw)
proc on /proc type proc (rw)
tmpfs on /dev/shm type tmpfs (rw)
devpts on /dev/pts type devpts (rw,gid=5,mode=620)
tmpfs on /tmp type tmpfs (rw,nosuid,nodev,mode=1777)
/dev/sysvg/lv_usr on /usr type ext3 (rw,noatime)
/dev/sysvg/lv_var on /var type reiserfs (rw)
/dev/sysvg/lv_local on /usr/local type ext3 (rw,noatime)
/dev/sysvg/lv_home on /home type reiserfs (rw,nosuid,nodev)
/dev/vg1/lv_misc on /misc type reiserfs (rw,nosuid,nodev)
/dev/vg1/lv_mirror on /mirror type reiserfs (rw,nosuid,nodev)
/dev/vg1/lv_hathi on /hathi type ext2 (ro)
capifs on /dev/capi type capifs (rw,mode=0666)
/proc/bus/usb on /proc/bus/usb type usbdevfs (rw)
AFS on /afs type afs (rw)
localhost:/var/lib/cfs/.cfsfs on /var/cfs type nfs (rw,port=3049,intr,nfsvers=2,addr=127.0.0.1)
binfmt_misc on /proc/sys/fs/binfmt_misc type binfmt_misc (rw)

- --- mdadm.conf
DEVICE partitions
ARRAY /dev/md0 level=raid1 num-devices=2 UUID=88cf7fb7:6fab12d7:b713c983:af6eaca5
MAILADDR root

- --- /proc/mdstat:
Personalities : [raid1] 
read_ahead 1024 sectors
md0 : active raid1 hdd3[0]
      57902464 blocks [2/1] [U_]
      
unused devices: <none>

- --- /proc/partitions:
major minor  #blocks  name     rio rmerge rsect ruse wio wmerge wsect wuse running use aveq

   9     0   57902464 md0 0 0 0 0 0 0 0 0 0 0 0
  58     1   32768000 lvmb 0 0 0 0 0 0 0 0 0 0 0
  58     2    6815744 lvmc 0 0 0 0 0 0 0 0 0 0 0
  58     3   32669696 lvmd 0 0 0 0 0 0 0 0 0 0 0
  58     4   36044800 lvme 0 0 0 0 0 0 0 0 0 0 0
  58     5    5242880 lvmf 0 0 0 0 0 0 0 0 0 0 0
  58     6     360448 lvmg 0 0 0 0 0 0 0 0 0 0 0
  58     9    1048576 lvmj 0 0 0 0 0 0 0 0 0 0 0
  22     0   58652496 hdc 288 2376 15528 2120 1127 7727 70854 8760 -20 582540 32154602
  22     1     249983 hdc1 11 269 560 80 0 0 0 0 0 60 80
  22     2     499968 hdc2 261 1815 14352 1890 1127 7727 70854 8760 0 4210 10670
  22     3   57902544 hdc3 11 269 560 130 0 0 0 0 0 110 130
  22    64  312571224 hdd 18594 122005 799916 834550 5509 12032 140304 934920 -8 581260 40123032
  22    65   78140159 hdd1 11 269 560 120 0 0 0 0 0 100 120
  22    66   78140160 hdd2 592 1287 7016 398410 2 0 16 0 0 3640 398410
  22    67   57906292 hdd3 17985 120423 792276 435960 5507 12032 140288 934920 0 91900 1370880
   3     0   39081623 hda 5080 21475 63598 21080 2274 12333 82500 16190 -21 582480 31594852
   3     1     249007 hda1 4803 19552 48710 18560 1403 4347 11598 7410 0 13900 25970
   3     2     498015 hda2 271 1881 14792 2450 871 7986 70902 8780 0 4270 11260

- --- initrd.img-2.4.32:

- --- /proc/modules:
raid1                  13264   1 (autoclean)
md                     58176   2 (autoclean) [raid1]

- --- /var/log/syslog:
Sep  4 09:45:09 ikki kernel: md: md driver 0.90.0 MAX_MD_DEVS=256, MD_SB_DISKS=27
Sep  4 09:45:09 ikki kernel: md: bind<hdd3,1>
Sep  4 09:45:09 ikki kernel: md: hdd3's event counter: 00000011
Sep  4 09:45:09 ikki kernel: md: raid1 personality registered as nr 3
Sep  4 09:45:09 ikki kernel: md: updating md0 RAID superblock on device
Sep  4 09:45:09 ikki kernel: md: hdd3 [events: 00000012]<6>(write) hdd3's sb offset: 57906176
Sep  4 09:45:09 ikki kernel: md: recovery thread got woken up ...
Sep  4 09:45:09 ikki kernel: md: recovery thread finished ...

- --- volume detail:
/dev/hdd3:
          Magic : a92b4efc
        Version : 00.90.00
           UUID : 88cf7fb7:6fab12d7:b713c983:af6eaca5
  Creation Time : Tue Jul 25 17:51:10 2006
     Raid Level : raid1
    Device Size : 57902464 (55.22 GiB 59.29 GB)
     Array Size : 57902464 (55.22 GiB 59.29 GB)
   Raid Devices : 2
  Total Devices : 1
Preferred Minor : 0

    Update Time : Mon Sep  4 09:44:43 2006
          State : active
 Active Devices : 1
Working Devices : 1
 Failed Devices : 0
  Spare Devices : 0
       Checksum : 955df62d - correct
         Events : 0.18


      Number   Major   Minor   RaidDevice State
this     0      22       67        0      active sync   /dev/hdd3

   0     0      22       67        0      active sync   /dev/hdd3
   1     1       0        0        1      faulty removed
- --


- -- System Information:
Debian Release: testing/unstable
  APT prefers testing
  APT policy: (700, 'testing'), (600, 'unstable'), (500, 'stable'), (60, 'experimental')
Architecture: i386 (i686)
Shell:  /bin/sh linked to /bin/dash
Kernel: Linux 2.4.32
Locale: LANG=de_DE, LC_CTYPE=de_DE (charmap=ISO-8859-1) (ignored: LC_ALL set to de_DE)

Versions of packages mdadm depends on:
ii  debconf [debconf-2.0]        1.5.3       Debian configuration management sy
ii  libc6                        2.3.6.ds1-4 GNU C Library: Shared libraries
ii  lsb-base                     3.1-14      Linux Standard Base 3.1 init scrip
ii  makedev                      2.3.1-82    creates device files in /dev

Versions of packages mdadm recommends:
ii  exim4-daemon-heavy [mail-tran 4.63-3     exim MTA (v4) daemon with extended

- -- debconf information:
* mdadm/autostart: true
* mdadm/initrdstart: none
* mdadm/warning:
* mdadm/start_daemon: true
* mdadm/mail_to: root
* mdadm/autocheck: false

- -- 
Klaus Ethgen                            http://www.ethgen.de/
pub  2048R/D1A4EDE5 2000-02-26 Klaus Ethgen <Klaus at Ethgen.de>
Fingerprint: D7 67 71 C4 99 A6 D4 FE  EA 40 30 57 3C 88 26 2B
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.5 (GNU/Linux)

iQEVAwUBRPveQp+OKpjRpO3lAQLopQf8DA22VaX1EXRCaK2IK8XAWiKolZW7KBWt
XwTqzznB3+dxbV8KXlWprwn7PdgMi9pwIQ4KlquxcDbc4XYTs4VvEo2fm4lSn4Ic
nocDESgPEBXwWLeQvVq1MVbsIlypayJDFNooU3qS8JpfxFFCpxJ/Rmd4HQa9RX0+
ZqpCWzJ/H9fwfYbEo/uMBUZOVVZyZ5oIUhWb+FGTRvAUZEDbFbtfyjoZZKn+eIkm
LD2u9K+OLqzQ2RWfZnak7/vwK/mP1Uc+z1ofZf96i0C3C+yR3FQaiYIe6iqhCjou
f/IlRsPtZ35Et70WodFeQ0D3DZ+diDBZh72TBqa5xsMdnVV519HOow==
=JTul
-----END PGP SIGNATURE-----




More information about the pkg-mdadm-devel mailing list