fsck question - help!

fsck question - help!

Post by TimK » Wed, 24 Jul 2002 00:40:24



The local power utility has an unfortunate habit of dropping
power to our building at random times.  Over the weekend the
most recent power disruption created an error on one of my
FreeBSD servers.

When the unit was powered back up the boot process stopped with
an error message.  This happens when it attempts to mount the
file systems.  The message reads:

ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
1395839; cn 167 tn 112 sn 111)
ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
1395839; cn 167 tn 112 sn 111)  status=59 error=40
ad0: DMA problem fallback to PIO mode
/dev/ad0s1e: CANNOT READ: BLK 67472
/dev/ad0s1e: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck
MANUALLY
THE FOLLOWING FILESYSTEM HAD AN UNEXPECTED INCONSISTENCY:
     /dev/ad0s1e: (/var)
Automatic filesystem check failed . . . help!
Enter full pathname of shell or RETURN for /bin/sh:

When I run fsck manually the same error text comes up and I
cannot get it to repair the damage to the file system.  No
questions about "do you want to fix this block?", just the same
old message (well, maybe a little different).

ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
1395839; cn 167 tn 112 sn 111)  status=59 error=40
ad0s1e: hard error reading fsbn 67473 (ad0s1 bn 67473; cn 8 tn
15 sn 0)  status=59 error=40
file system still dirty
rerun fsck

I know that this is why there are backups, but is there any
other way around this problem, has the hard drive shot craps
(snake-eyes or box cars, your choice) ?  I am running 4.5 with
softupdates enabled.

Any suggestions are welcome.

Like the text says - help!!

 
 
 

fsck question - help!

Post by Lars Tunkran » Wed, 24 Jul 2002 01:18:54



> The local power utility has an unfortunate habit of dropping
> power to our building at random times.  Over the weekend the
> most recent power disruption created an error on one of my
> FreeBSD servers.

> When the unit was powered back up the boot process stopped with
> an error message.  This happens when it attempts to mount the
> file systems.  The message reads:

> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40
> ad0: DMA problem fallback to PIO mode
> /dev/ad0s1e: CANNOT READ: BLK 67472
> /dev/ad0s1e: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck
> MANUALLY
> THE FOLLOWING FILESYSTEM HAD AN UNEXPECTED INCONSISTENCY:
>      /dev/ad0s1e: (/var)
> Automatic filesystem check failed . . . help!
> Enter full pathname of shell or RETURN for /bin/sh:

> When I run fsck manually the same error text comes up and I
> cannot get it to repair the damage to the file system.  No
> questions about "do you want to fix this block?", just the same
> old message (well, maybe a little different).

> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40
> ad0s1e: hard error reading fsbn 67473 (ad0s1 bn 67473; cn 8 tn
> 15 sn 0)  status=59 error=40
> file system still dirty
> rerun fsck

> I know that this is why there are backups, but is there any
> other way around this problem, has the hard drive shot craps
> (snake-eyes or box cars, your choice) ?  I am running 4.5 with
> softupdates enabled.

    I can not answer your question, However I had a Hard disk error
On another Unix system to weeks ago - an old sVr4 system - so :

   On SCSI disks there is  a couple of spare sectors per cylinder that
   a Modern SCSI driver is supposed to use to reallocate a failed sector
   automatically. Ofcourse the Spare sectors MAY have been alredy used
   in your machine.

    On my old machine we had two utilities :

     readblk, that did a single read pass over the whole surface of a
     disk and noted any bad blocks.

      findblk, that was an interactive badblock reallocation utility
      It tried to reallocate on a block by block basis the bad blocks
      and asked you for permission for every move.

     Using these small utilities I Conviced a disk to live a litle while
     longer. Is still running. Manufacture date 1994.

   NOW;  for the experts:

      How does bad block/sector realloation work in FreeBSD ?
         Does it work on both  IDE and SCSI disks ?

     Regards
   //LArs

 
 
 

fsck question - help!

Post by Steven G. Kar » Wed, 24 Jul 2002 03:19:37




Quote:

> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40
> ad0: DMA problem fallback to PIO mode
> /dev/ad0s1e: CANNOT READ: BLK 67472
> /dev/ad0s1e: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck
> MANUALLY
> THE FOLLOWING FILESYSTEM HAD AN UNEXPECTED INCONSISTENCY:
>      /dev/ad0s1e: (/var)
> Automatic filesystem check failed . . . help!
> Enter full pathname of shell or RETURN for /bin/sh:

This isn't pretty.  I believe the local power company
may have toasted the hard drive.  

Quote:> When I run fsck manually the same error text comes up and I
> cannot get it to repair the damage to the file system.  No
> questions about "do you want to fix this block?", just the same
> old message (well, maybe a little different).

What was the exact fsck command you ran?

Have you "fsck -y" and "fsck -p"?
You may need to use an alternate superblock, so
try "fsck_ffs -b 32 -y".

If all else fails and you have info on that partition
that you absolutely need.  You can forcefully mount the
dirty partition and try to copy as much info from ad
you can.  This can be dangerous, so use it with caution.

--
Steve
http://troutmask.apl.washington.edu/~kargl/

 
 
 

fsck question - help!

Post by p.. » Wed, 24 Jul 2002 06:30:28



> The local power utility has an unfortunate habit of dropping
> power to our building at random times.  Over the weekend the
> most recent power disruption created an error on one of my
> FreeBSD servers.
> When the unit was powered back up the boot process stopped with
> an error message.  This happens when it attempts to mount the
> file systems.  The message reads:
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40
> ad0: DMA problem fallback to PIO mode
> /dev/ad0s1e: CANNOT READ: BLK 67472
> /dev/ad0s1e: UNEXPECTED SOFT UPDATE INCONSISTENCY; RUN fsck
> MANUALLY
> THE FOLLOWING FILESYSTEM HAD AN UNEXPECTED INCONSISTENCY:
>      /dev/ad0s1e: (/var)
> Automatic filesystem check failed . . . help!
> Enter full pathname of shell or RETURN for /bin/sh:
> When I run fsck manually the same error text comes up and I
> cannot get it to repair the damage to the file system.  No
> questions about "do you want to fix this block?", just the same
> old message (well, maybe a little different).
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40
> ad0s1e: hard error reading fsbn 67473 (ad0s1 bn 67473; cn 8 tn
> 15 sn 0)  status=59 error=40
> file system still dirty
> rerun fsck
> I know that this is why there are backups, but is there any
> other way around this problem, has the hard drive shot craps
> (snake-eyes or box cars, your choice) ?  I am running 4.5 with
> softupdates enabled.

Read the above " hard error reading fsbn 1395839"

The disk has an blind spot where fsck would like to write. The place
that was used has become bad. fsck does not do bad-block remapping.

Your best bet is to allocate another partition, maybe on a new disk,
mount /var read-only, copy everything to your new /var and then forget
the old partition until you can replace the drive.

This is why i always keeps a spare partition on my disks.

Quote:> Any suggestions are welcome.
> Like the text says - help!!

--
Peter H?kanson        
        IPSec  Sverige      ( At Gothenburg Riverside )
           Sorry about my e-mail address, but i'm trying to keep spam out,
           remove "icke-reklam" if you feel for mailing me. Thanx.
 
 
 

fsck question - help!

Post by mic.. » Wed, 24 Jul 2002 17:01:22



> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)
> ad0s1e: hard error reading fsbn 1395839 of 67472-67473 (ad0s1 bn
> 1395839; cn 167 tn 112 sn 111)  status=59 error=40

So you have a real bad block on the disk. In principle bad blocks are
automatically remapped, this means that you may have to buy another
disk.

Quote:> When I run fsck manually the same error text comes up and I
> cannot get it to repair the damage to the file system.  No
> questions about "do you want to fix this block?", just the same
> old message (well, maybe a little different).

Do you answer YES anytime fsck proposes to change something? If not
don't be surprised if fsck fails.
By the way you should be able to mount the disk read only without
fsck ing it and copy it entirely to another disk.

--
Michel Talon

 
 
 

1. fsck fsck fsck say I

[[[Ouch.  Sorry.  Ouch.  The first edition of this text lost the first
word of its Subject.  Fixed now]]]

I remember I first learned of `fsck -f -y /` after `umount /` in the
thread:

Newsgroups: comp.unix.shell
Subject: fsck fsck fsck say I
http://groups.google.com/groups?threadm=2695edf1.0312021221.72f242a3%...

Speaking again now of fsck'ing on demand, eventually I also noticed,
...

My `man -k fsck` does not yet, but could/ should, point to:

--- http://www.google.com/search?q=linux+man+shutdown
--- http://www.die.net/doc/linux/man/man8/shutdown.8.html
...
-f
Skip fsck on reboot.
-F
Force fsck on reboot.
...
---

Pat LaVarre

2. Best Linux distribution?

3. xntpdc passwd ??

4. help pls - fsck fail on boot; fixed, fsck still fails on boot

5. Atlanta UNIX Admin.

6. Help! FSCK question

7. Help crashing SCSI

8. HELP: fsck question

9. help: / /usr fsck question

10. fsck fsck say I

11. fsck question

12. fsck and redhat question