Bugzilla – Bug 731230
cannot boot machine with md raid disks.
Last modified: 2017-08-11 18:44:34 UTC
User-Agent: Mozilla/5.0 (X11; Linux x86_64; rv:7.0.1) Gecko/20100101 Firefox/7.0.1 under opensuse 11.4 we were able to mount a raid file system during the system boot. under 12.4, this causes the booting to fail. The likely cause of this failure is some missing dependency in the startup order. The value of the sixth parameter of the fstab entry had been 2. But apparently this is no longer good enough. We can get around the problem by setting the mount parameters for the raid partition to be noauto and having the raid partition mounted after the boot finishes by using a cron job. This is a hack, but it works. Reproducible: Always Steps to Reproduce: 1. create a raid file system that you want to mount during boot time. 2. reboot your machine. 3. Actual Results: the system goes into single user mode because the file system cannot be properly fscked or mounted. mounting it from single user mode followed by init 5 works.
Does it work if you boot with SysV init? (you can toggle at boot prompt with F5) Please attach /etc/fstab How does the mounting fail? Does it hang or just does not mount the partition?
Hi, same problems here, have several system with different md devices like this /dev/md0 swap swap defaults 0 0 /dev/md1 / ext4 acl,user_xattr 1 1 /dev/md2 /usr/local ext4 defaults 1 2 Everythink worked well for openSuSE 11.4. I made a "zypper dup" with the new 12.1 repositories. Now the boot process hangs but can be continued manuelly with ctrl-D. I think systemd tries to mount /dev/md2 before it has been enabled: <6>[ 0.000000] Initializing cgroup subsys cpuset <6>[ 0.000000] Initializing cgroup subsys cpu <5>[ 0.000000] Linux version 3.1.0-1.2-desktop (geeko@buildhost) (gcc version 4.6.2 (SUSE Linux) ) #1 SMP PREEMPT Thu Nov 3 14:45:45 UTC 2011 (187dde0) <6>[ 0.000000] Command line: root=/dev/md1 splash=verbose quiet vga=0x31a ... <6>[ 10.170018] md: bind<sda2> <6>[ 10.292817] md: bind<sda4> <6>[ 10.774111] device-mapper: uevent: version 1.0.3 <6>[ 10.774200] device-mapper: ioctl: 4.21.0-ioctl (2011-07-06) initialised: dm-devel@redhat.com <30>[ 10.794925] swapon[762]: swapon: /dev/md0: read swap header failed: Das Argument ist ungültig <29>[ 10.795248] systemd[1]: dev-md0.swap swap process exited, code=exited status=255 <29>[ 10.809986] systemd[1]: Unit dev-md0.swap entered failed state. <6>[ 10.838103] md: bind<sdb4> <30>[ 10.864658] boot.lvm[629]: Reading all physical volumes. This may take a while... <30>[ 10.864814] boot.lvm[629]: Activating LVM volume groups... <6>[ 10.904426] md/raid1:md2: active with 2 out of 2 mirrors <6>[ 10.904440] md2: detected capacity change from 0 to 444600942592 <30>[ 10.912624] boot.md[620]: Starting MD RAID mdadm: /dev/md0 is already in use. <6>[ 10.947558] md2: unknown partition table <30>[ 11.152105] systemd-fsck[768]: fsck.ext4: Das Argument ist ungültig beim Versuch, /dev/md2 zu öffnen <30>[ 11.152602] systemd-fsck[768]: /dev/md2: <27>[ 11.152606] systemd-fsck[768]: fsck failed with error code 8. <28>[ 11.152609] systemd-fsck[768]: Ignoring error. <30>[ 11.152612] systemd-fsck[768]: SuperBlock ist unlesbar bzw. beschreibt kein gültiges ext2 <30>[ 11.153080] systemd-fsck[450]: Dateisystem. Wenn Gerät gültig ist und ein ext2 <30>[ 11.153336] systemd-fsck[450]: Dateisystem (kein swap oder ufs usw.) enthält, dann ist der SuperBlock <30>[ 11.153658] systemd-fsck[450]: beschädigt, und sie könnten e2fsck mit einem anderen SuperBlock: <30>[ 11.154032] systemd-fsck[450]: e2fsck -b 8193 <Gerät> <30>[ 11.231812] boot.lvm[629]: No volume groups found <30>[ 11.232332] boot.lvm[629]: ..done <6>[ 11.405439] md: bind<sdb2> <6>[ 11.526691] md/raid1:md0: active with 2 out of 2 mirrors <6>[ 11.526705] md0: detected capacity change from 0 to 9179299840 <6>[ 11.610110] EXT4-fs (md1): re-mounted. Opts: acl,user_xattr <6>[ 11.610283] md0: <30>[ 11.655265] mount[798]: mount: /dev/md2 ist bereits eingehängt oder /usr/local wird gerade benutzt <29>[ 11.655636] systemd[1]: usr-local.mount mount process exited, code=exited status=32 <29>[ 11.675220] systemd[1]: Job klog.service/start failed with result 'dependency'. <29>[ 11.675227] systemd[1]: Job remote-fs.target/start failed with result 'dependency'. <29>[ 11.675233] systemd[1]: Job local-fs.target/start failed with result 'dependency'. <30>[ 11.675236] systemd[1]: Triggering OnFailure= dependencies of local-fs.target. <29>[ 11.675690] systemd[1]: Unit usr-local.mount entered failed state. <29>[ 12.246362] systemd[1]: md.service: control process exited, code=exited status=3 <29>[ 12.274239] systemd[1]: Unit md.service entered failed state. <30>[ 12.282953] boot.md[844]: Not shutting down MD RAID - reboot/halt scripts do this...missing <30>[ 12.293363] systemd[1]: Startup finished in 4s 842ms 507us (kernel) + 7s 329ms 795us (userspace) = 12s 172ms 302us. <30>[ 21.087131] udevd[860]: starting version 173 <30>[ 21.107736] boot.corefile[864]: Setting core dump file name <6>[ 21.117180] Adding 8964156k swap on /dev/md0. Priority:0 extents:1 across:8964156k <30>[ 21.169633] boot.localnet[926]: Using boot-specified hostname 'opt80' <30>[ 21.170340] boot.localnet[926]: Setting up hostname 'opt80'..done <30>[ 21.171719] boot.localnet[926]: Setting up loopback interface RTNETLINK answers: File exists <30>[ 21.172684] boot.localnet[926]: ..done <14>[ 21.276422] mtp-probe[1044]: checking bus 1, device 3: "/sys/devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3" <14>[ 21.277233] mtp-probe[1044]: bus: 1, device: 3 was not an MTP device <14>[ 21.283651] mtp-probe[1050]: checking bus 2, device 3: "/sys/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.7" <14>[ 21.285257] mtp-probe[1050]: bus: 2, device: 3 was not an MTP device <14>[ 21.286139] mtp-probe[1054]: checking bus 2, device 4: "/sys/devices/pci0000:00/0000:00:1d.0/usb2/2-1/2-1.8" <30>[ 21.293286] systemd-fsck[921]: /dev/md2: sauber, 27884/27140096 Dateien, 55382685/108545152 Blöcke <14>[ 21.297437] mtp-probe[1054]: bus: 2, device: 4 was not an MTP device <30>[ 21.304346] boot.lvm[1056]: Waiting for udev to settle... <13>[ 21.320311] ifup[1139]: Service network not started and mode 'auto' -> skipping <30>[ 21.665124] boot.md[928]: Starting MD RAID ..done <6>[ 21.715251] EXT4-fs (md1): re-mounted. Opts: acl,user_xattr <30>[ 21.839334] boot.lvm[1056]: Scanning for LVM volume groups... <6>[ 21.905398] EXT4-fs (md2): mounted filesystem with ordered data mode. Opts: (null) Sys V init completely hangs and requires a ctrl-C to continue: <6>[ 3.831689] EXT4-fs (md1): mounted filesystem with ordered data mode. Opts: acl,user_xattr <6>[ 3.919706] EXT4-fs (md1): re-mounted. Opts: acl,user_xattr <30>[ 163.765329] udevd[439]: starting version 173 <6>[ 164.224244] input: Power Button as /devices/LNXSYSTM:00/device:00/PNP0C0C:00/input/input2 <6>[ 164.224290] ACPI: Power Button [PWRB] <6>[ 164.224325] input: Power Button as /devices/LNXSYSTM:00/LNXPWRBN:00/input/input3 <6>[ 164.224383] ACPI: Power Button [PWRF] <6>[ 164.248536] input: PC Speaker as /devices/platform/pcspkr/input/input4
hi, Same problem here: during boot sequence, mounting / or /home fails with superblock problem, sending me to recovery console. Exiting console with ^D leads to a successful boot. Here's my fstab: % cat /etc/fstab /dev/disk/by-id/ata-SAMSUNG_HD154UI_XXX-part4 swap swap defaults 0 0 /dev/disk/by-id/ata-SAMSUNG_HD154UI_XXX-part4 swap swap defaults 0 0 /dev/md1 / ext4 acl,user_xattr 1 1 /dev/md0 /boot ext4 acl,user_xattr 1 2 /dev/md2 /home ext4 acl,user_xattr 1 2 proc /proc proc defaults 0 0 sysfs /sys sysfs noauto 0 0 debugfs /sys/kernel/debug debugfs noauto 0 0 devpts /dev/pts devpts mode=0620,gid=5 0 0
Andreas Jaeger, it is a holiday week in the US and I am out of town and so I cannot take that machine down to get the info until next week. However, the other two bug reports look very much like my failure, in particular, the fstab entry for /dev/md2 is very similar to mine. I even think that the console messages are similar except of course, mine are in english. The one difference between mine and oliverp's system is that my root file system is a conventional partition, and the md file system that i am having trouble with is where we have our svn server so it is mounted with a 2 in column 6. The one thing that might be relevant is that my machine is a 4 core machine. it could be that with fewer cores, the problem is not visible. I will upload additional files if necessary next week, but it looks like this problem may not be hard to reproduce. Kenny
(In reply to comment #4) > > The one thing that might be relevant is that my machine is a 4 core machine. > it could be that with fewer cores, the problem is not visible. > I have 4 cores here as well (Athlon 2 X4). Same bug here : https://bugzilla.novell.com/show_bug.cgi?id=731135
Same bug here with systemd, superblock error mounting md0 at boot. With an 4 cores processor too (qcore 6600, 64bits, 12.1). With traditional sysvinit seems to work fine. My fstab: /dev/system/root / ext4 acl,user_xattr 1 1 /dev/md0 /boot ext4 acl,user_xattr 1 2 /dev/system/swap swap swap defaults 0 0 /dev/system/var /var ext4 acl,user_xattr 1 2 /dev/system/home /home ext4 acl,user_xattr 1 2 /dev/system/usrlocal /usr/local ext4 acl,user_xattr 1 2 proc /proc proc defaults 0 0 sysfs /sys sysfs noauto 0 0 debugfs /sys/kernel/debug debugfs noauto 0 0 usbfs /proc/bus/usb usbfs noauto 0 0 devpts /dev/pts devpts mode=0620,gid=5 0 0 nfsserver:/ /srv/share nfs4 defaults,bg 0 0 /dev/md0 => 2 disks raid 1, version 0.90. I tried version 1.2 too, with same results. LVM is on top of 2 more raids 1, first of them is in the same disk than md0. They work fine. The workarround,putting md0 like noauto, and mounting manually in boot.local (i.e.), works.
I believe that bug 731135 is for this same problem. As far as I can tell, the essence of the problem is that systemd doesn't have an explicit md.service description and associated target files (see /lib/systemd/system/localfs.service for comparison), and doesn't know about the dependency that boot.md (or its systemd equivalent) *must* complete before boot.localfs (or equivalent) is started. In the SysV init world, boot.md runs before boot.localfs because of the sequence numbers assigned to them in "/etc/init.d/boot.d". The SysV init scripts also contain LSB comments that are supposed to document service startup and shutdown dependencies. I think the LSB comments for boot.localfs are incorrect: they seem to say that boot.localfs runs before boot.md, when it must be the other way around.
*** Bug 731135 has been marked as a duplicate of this bug. ***
Andreas, Olivier, I am back from vacation. Is there any other information that I could provide to help you with this bug. It looks like others have supplied the logs you asked for. Kenny
Thinking about this problem some more, what's really desired, and what systemd should be designed to support, is a finer-grained system that tracks dependencies on a per-filesystem basis. For example, ordinary local filesystems could be mounted in parallel with each other, and in parallel with the mdadm instance that is building the RAID devices (/dev/md0, etc.). Once a new raid device is available, an event of some sort would be generated so systemd could then attempt to mount the filesystem on the new device. This could be viewed as an extension of mount point ordering -- there might be filesystems that with mount points that lie on top of the RAID-based filesystem. It might also provide a better system administration experience in case of partial drive or filesystem failure. A compromise might be to split localfs mounts into two phases: one that mounts filesystems that don't depend upon md drives, and one that mounts filesystems that do depend on md drives. Ultimately, I imagine, all this complexity will be hidden in the filesystems themselves, such as with the RAID support integrated into BTRFS. In the short term, I'd be satisfied to have openSUSE 12.1's systemd-based initialization build all the md devices, then mount the local filesystems, just like the SysV init package does.
could people test package from home:fcrozat:systemd / systemd ?
Hi, This solved it : sudo zypper ar "http://download.opensuse.org/repositories/home:/fcrozat:/systemd/openSUSE_12.1/home:fcrozat:systemd.repo" sudo zypper dup Thanks a lot. Regards, Olivier
/dev/md0 is a raid1 version 0.90 (default is 1.0) Tested: Packages installed: systemd|37-297.1|x86_64||home_fcrozat_systemd systemd-presets-branding-openSUSE|0.1.0-29.1|noarch systemd-sysvinit|37-297.1|x86_64||home_fcrozat_systemd I rebooted a lot, and my system boots every time. But, in 3 times, checking with "systemctl --failed" a md.service error apears: # systemctl status md.service md.service - LSB: Multiple Device RAID Loaded: loaded (/etc/init.d/boot.md) Active: failed since Tue, 29 Nov 2011 21:52:32 +0100; 1min 51s ago Process: 706 ExecStart=/etc/init.d/boot.md start (code=exited, status=1/FAILURE) CGroup: name=systemd:/system/md.service despite that, /boot remains mounted correctly, and no more "fsck" and "SuperBlock" errors: [ 11.970454] systemd[1]: md.service: control process exited, code=exited status=1 [ 11.997939] systemd[1]: Unit md.service entered failed state. [ 12.081093] md: bind<sda1> [ 12.178361] md/raid1:md0: active with 2 out of 2 mirrors [ 12.185430] md0: detected capacity change from 0 to 1003356160 [ 12.225429] md0: unknown partition table [ 12.856998] systemd-fsck[926]: boot: limpio, 43/61312 ficheros, 15060/244960 bloques [ 12.870665] systemd-fsck[929]: var: limpio, 1845/1966080 ficheros, 717746/7864320 bloques [ 12.884267] systemd-fsck[932]: home: limpio, 423283/68911104 ficheros, 164044109/275633152 bloques (comprobación después de 3 montajes) [ 12.904727] systemd-fsck[935]: usrlocal: limpio, 290/655360 ficheros, 86311/2621440 bloques (comprobación en el siguiente montaje) [ 13.082309] EXT4-fs (dm-2): mounted filesystem with ordered data mode. Opts: acl,user_xattr [ 13.103274] EXT4-fs (md0): mounted filesystem with ordered data mode. Opts: acl,user_xattr [ 13.135768] EXT4-fs (dm-1): mounted filesystem with ordered data mode. Opts: acl,user_xattr [ 13.273474] EXT4-fs (dm-4): mounted filesystem with ordered data mode. Opts: acl,user_xattr
ok, will be handled by maintenance update of systemd *** This bug has been marked as a duplicate of bug 724912 ***
Please, could some one of you, reporters or assignees, add an entry for this issue in the "openSUSE:Most annoying bugs 12.1" wiki page? http://en.opensuse.org/openSUSE:Most_annoying_bugs_12.1
while booting from HDD it is giving that Failed Features : boot.localfs Skipped features : boot.md
12.1 is out of support scope and currently, we test MD raid boots in openQA so they should be working just fine. Please open a new bug if the problem is still present on current Leap or Tumbleweed releases.