[systemd-devel] fsck checking: passno interpretation problem

Mariusz Bialonczyk manio at skyboo.net
Fri Jul 29 16:11:03 PDT 2011


On 05/17/2011 10:50 AM, Mariusz Bialonczyk wrote:
> Hi
> When I was switching to systemd on another debian box I've discovered
> a problem related with passno field in fstab.
> I have several md partitions (I am using kernel autoassembly for md,
> and I am not using initrd).
> I have systemd v.25-2 (debian experimental).
> 
> I have the following entries in fstab for my partitions:
> UUID=f73bcffa-5b89-4f1d-a296-5f0e6ebf4673 /               ext4    errors=remount-ro 0       1
> UUID=33a13f77-6e5a-4c1e-b959-d891b0a77488 /home           ext4    defaults        0       2
> UUID=45e28f49-1822-4b32-b8a4-2a9ac1bf543e /usr/src        ext4    defaults        0       2
> UUID=8fe78553-ad5e-47f1-8630-bf1fa5d7b608 /var            ext4    defaults        0       2
> UUID=af8133a0-e9a5-4fcf-81e4-c87c2bb50ac8 none            swap    sw              0       0
> all of this UUIDs are /dev/md devices - and not counting the mirroring
> disk, all of theese partition is on one physical sata disk.
> 
> According to manual all is fine here:
> The  sixth field, (fs_passno), is used by the fsck(8) program to deter-
> mine the order in which filesystem checks are done at reboot time.  The
> root  filesystem  should  be specified with a fs_passno of 1, and other
> filesystems should have a fs_passno of 2.  Filesystems within  a  drive
> will  be checked sequentially, but filesystems on different drives will
> be checked at the same time to utilize  parallelism  available  in  the
> hardware.   If  the sixth field is not present or zero, a value of zero
> is returned and fsck will assume that the filesystem does not  need  to
> be checked.
> 
> But this cause a problem to systemd when booting:
> systemd-fsck[1346]: /dev/md0: clean, 86170/458752 files, 954201/1834480 blocks
> systemd-fsck[3078]: /dev/md2: clean, 312227/1966080 files, 2639961/7864304 blocks (check in 2 mounts)
> systemd-fsck[3091]: fsck.ext4: Device or resource busy while trying to open /dev/md4
> systemd-fsck[3091]: Filesystem mounted or opened exclusively by another program?
> systemd-fsck[3087]: /dev/md1 has been mounted 23 times without being checked, check forced.
> systemd-fsck[3087]: /dev/md1: 8010/983040 files (1.2% non-contiguous), 2091221/3932144 blocks
> Welcome to emergency mode. Use "systemctl default" or ^D to activate default mode.
> Give root password for maintenance
> (or type Control-D to continue):
> 
> I am not 100% sure but if I remind correctly, the "busy one" partition
> was random within reboots.
> 
> After conversation with Kay Sievers I've set my fstab entries to this:
> UUID=f73bcffa-5b89-4f1d-a296-5f0e6ebf4673 /               ext4    errors=remount-ro 0       1
> UUID=33a13f77-6e5a-4c1e-b959-d891b0a77488 /home           ext4    defaults        0       2
> UUID=45e28f49-1822-4b32-b8a4-2a9ac1bf543e /usr/src        ext4    defaults        0       3
> UUID=8fe78553-ad5e-47f1-8630-bf1fa5d7b608 /var            ext4    defaults        0       4
> UUID=af8133a0-e9a5-4fcf-81e4-c87c2bb50ac8 none            swap    sw              0       0
> 
> I changed the passno fields to 1,2,3,4,0 from 1,2,2,2,0 and this fixes
> the problem and my system is now bootable:
> systemd-fsck[1334]: /dev/md0: clean, 86178/458752 files, 954201/1834480 blocks
> systemd-fsck[3025]: /dev/md4: clean, 2675040/57581568 files, 199580895/230296816 blocks
> systemd-fsck[3109]: /dev/md2 has been mounted 28 times without being checked, check forced.
> systemd-fsck[3109]: /dev/md2: 312227/1966080 files (0.4% non-contiguous), 2639961/7864304 blocks
> systemd-fsck[3190]: /dev/md1: clean, 8033/983040 files, 2067703/3932144 blocks
> Setting console screen modes.
> Skipping font and keymap setup (handled by console-setup).
> ... and login prompt after a while
> 
> So the problem is somewhere with passno interpretation (according to
> manual it should work, but it doesn't).
> 
> regards,

Hello again
I want to bring some new light on this issue. It seems that passno doesn't solve the
"busy device" issue (I tested it on debian's systemd v25-2 and also 29-1).


More information about the systemd-devel mailing list