UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck MANUALLY

Clint Olsen clint.olsen at gmail.com
Mon Sep 22 00:16:45 UTC 2008


On Sep 21, Jeremy Chadwick wrote:
> Re-adding mailing list to the CC list.
> 
> No, I don't think that is the case, assuming the filesystems are UFS2
> and are using softupdates.  When booting multi-user, fsck is run in the
> background, meaning the system is fully up + usable even before the fsck
> has started.

The last time things crashed hard, the boot sequence was halted in order to
run fsck.
 
> Consider using background_fsck="no" in /etc/rc.conf if you prefer the
> old behaviour.  Otherwise, boot single-user then do the fsck.

Yes, I'll do this.
 
> You could also consider using clri(8) to clear the inode (190).  Do this
> in single-user while the filesystem is not mounted.  After using clri,
> run fsck a couple times.

Ok, thanks.
 
> Also, are there any kernel messages about ATA/SCSI disk errors or other
> anomalies?

None.  In fact smartctl will not do anything now.  It just prints out the
quick banner message and exits immediately with no error.

-Clint

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.



More information about the freebsd-stable mailing list