Inconsistency in root partition size

Adam Vande More amvandemore at gmail.com
Mon Sep 7 03:08:53 UTC 2009


 Sun, Sep 6, 2009 at 9:24 PM, jaymax <jaymax36 at gmail.com> wrote:

>
>
>
> Mel Flynn-2 wrote:
> >
> >
> > No, single user mode. Root partition in single user mode can be fsck'd
> and
> > repaired if mounted ro in single user. The system does fsck -p by
> default,
> > which skips partitions marked clean. Since you can shutdown cleanly,
> > nothing
> > will happen.
> > Have a look at /etc/defaults/rc.conf for setting different behaviors by
> > overriding the defaults in /etc/rc.conf.
> >
>
> In single user mode I ran fsck -vy /
> there were a few Salvages along the way
> Rebooted but the condition remained unchanged.
> rebooted again in single user mode ran fsck -vy /dev/ad0s1a
> everything seemed clear - no salvage operation
> Normal reboot, but discrepancy still persits
>
> Hope we have not reached a checkmate!
>
> from the logs the only things discernable, were
> dmesg =>
>
>
> > ad0: 39083MB <Maxtor 5T040H4 TAH71DP0> at ata0-master UDMA100
> > ad1: 76345MB <MAXTOR 6L080J4 A93.0500> at ata0-slave UDMA100
> > acd0: CDROM <CD-ROM Drive/G6D/M1.30> at ata1-master UDMA33
> > acd1: DVDR <NEC DVD RW ND-3550A/1.05> at ata1-slave UDMA33
> > ad4: 381554MB <Seagate ST3400633A 3.AAH> at ata2-master UDMA100
> > Trying to mount root from ufs:/dev/ad0s1a
> > WARNING: /usr was not properly dismounted
> > WARNING: /disk02 was not properly dismounted
> > WARNING: /disk03 was not properly dismounted
> >
>
> from /var/all.log the record for today
>
> Earlier mount attempt
>
> # grep -in ad0s1a all.log
> 143:Sep  6 02:23:41 mach_1 kernel: Trying to mount root from
> ufs:/dev/ad0s1a
> 158:Sep  6 02:23:41 mach_1 kernel: /dev/ad0s1a: DEFER FOR BACKGROUND
> CHECKING
> 472:Sep  6 02:25:02 mach_1 fsck: /dev/ad0s1a: CANNOT CREATE SNAPSHOT
> //.snap/fsck_snapshot: No space left on device
> 474:Sep  6 02:25:02 mach_1 fsck: /dev/ad0s1a: UNEXPECTED INCONSISTENCY; RUN
> fsck MANUALLY.
> 985:Sep  6 03:11:21 mach_1 kernel: Trying to mount root from
> ufs:/dev/ad0s1a
> 1000:Sep  6 03:11:21 mach_1 kernel: /dev/ad0s1a: DEFER FOR BACKGROUND
> CHECKING
> 1315:Sep  6 03:13:08 mach_1 fsck: /dev/ad0s1a: CANNOT CREATE SNAPSHOT
> //.snap/fsck_snapshot: No space left on device
> 1317:Sep  6 03:13:08 mach_1 fsck: /dev/ad0s1a: UNEXPECTED INCONSISTENCY;
> RUN
> fsck MANUALLY.
>
>
> last mount
>
>
> 6311:Sep  6 17:39:33 mach_1 kernel: Trying to mount root from
> ufs:/dev/ad0s1a
> 6325:Sep  6 17:39:33 mach_1 kernel: /dev/ad0s1a: FILE SYSTEM CLEAN;
> SKIPPING
> CHECKS
> 6326:Sep  6 17:39:33 mach_1 kernel: /dev/ad0s1a: clean, 1982 free (1966
> frags, 2 blocks, 0.8% fragmentation)
> 6772:Sep  6 17:51:19 mach_1 kernel: Trying to mount root from
> ufs:/dev/ad0s1a
> 6786:Sep  6 17:51:19 mach_1 kernel: /dev/ad0s1a: FILE SYSTEM CLEAN;
> SKIPPING
> CHECKS
> 6787:Sep  6 17:51:19 mach_1 kernel: /dev/ad0s1a: clean, 1968 free (1968
> frags, 0 blocks, 0.8% fragmentation)
>
>
> and
> df -k =>
> df -k
> Filesystem  1K-blocks      Used    Avail Capacity  Mounted on
> /dev/ad0s1a    507630    503694   -36674   108%    /
>
> Don't know if the above can show anything
> Could this be a kernel issue, tuning etc?
>
> Thanks!
> --
>
> ufs by default keeps a certain portion in reserve for use by root.  8% is
the standard amount I believe so that capacity reading is technically
valid.  On rare occasion, I've had to run fsck multiple times.  you may wish
to try this also, w/ no reboot in between.


-- 
Adam Vande More


More information about the freebsd-questions mailing list