[gentoo-user] How do I show list of bad blocks on a disk?

[gentoo-user] How do I show list of bad blocks on a disk?

[Nikos Chantziaras]

Before leaving home, I started an fsck.ext4 on a filesystem (500GB) that 
resides on a disk that I suspect is damaged:

   fsck.ext4 -c -c -f /dev/sdb1

When I came back 10 hours later, it was still checking.  After 2 hours 
more (so it took 12 hours total) it finally finished.  The output was:

   e2fsck 1.41.14 (22-Dec-2010)
   Checking for bad blocks (non-destructive read-write test)
   Testing with random pattern: done
   Extra: Updating bad block inode.
   Pass 1: Checking inodes, blocks, and sizes
   Pass 2: Checking directory structure
   Pass 3: Checking directory connectivity
   Pass 4: Checking reference counts
   Pass 5: Checking group summary information

   Extra: ***** FILE SYSTEM WAS MODIFIED *****
   Extra: 11/30531584 files (0.0% non-contiguous),
   1966902/122096638 blocks

I'm not sure how to read this.  Were there any bad blocks or not?  Is 
there a way to query the filesystem for the now known bad blocks?  (The 
"Updating bad block inode." message suggests that such a list is stored 
directly inside the filesystem.)

Re: [gentoo-user] How do I show list of bad blocks on a disk?

[Alex Schuster]

Nikos Chantziaras writes:

> Before leaving home, I started an fsck.ext4 on a filesystem (500GB) that 
> resides on a disk that I suspect is damaged:
> 
>    fsck.ext4 -c -c -f /dev/sdb1
> 
> When I came back 10 hours later, it was still checking.  After 2 hours 
> more (so it took 12 hours total) it finally finished.  The output was

Anything about erros in dmesg or syslog?

>    e2fsck 1.41.14 (22-Dec-2010)
>    Checking for bad blocks (non-destructive read-write test)
>    Testing with random pattern: done
>    Extra: Updating bad block inode.
>    Pass 1: Checking inodes, blocks, and sizes
>    Pass 2: Checking directory structure
>    Pass 3: Checking directory connectivity
>    Pass 4: Checking reference counts
>    Pass 5: Checking group summary information
> 
>    Extra: ***** FILE SYSTEM WAS MODIFIED *****
>    Extra: 11/30531584 files (0.0% non-contiguous),
>    1966902/122096638 blocks
> 
> I'm not sure how to read this.  Were there any bad blocks or not?  Is 
> there a way to query the filesystem for the now known bad blocks?  (The 
> "Updating bad block inode." message suggests that such a list is stored 
> directly inside the filesystem.)

dumpe2fs -b /dev/sdb1 probably also works for ext4.

bablocks /dev/sdb2 will do a read-only check of the whole partiton for
bad blocks. Use option -n for a non-destructive write mode.
I qalso like to add options -s and -v to see the progress. I redirect
the output into a file then, because output of progress and bad blocks
will overlap: badblocks -sv /dev/sdb1 > sdb1.bad
See man badblocks for more information.

	Wonko

[gentoo-user] Re: How do I show list of bad blocks on a disk?

[Nikos Chantziaras]

On 03/06/2011 07:25 PM, Alex Schuster wrote:
> Nikos Chantziaras writes:
>
>> Before leaving home, I started an fsck.ext4 on a filesystem (500GB) that
>> resides on a disk that I suspect is damaged:
>>
>>     fsck.ext4 -c -c -f /dev/sdb1
>>
>> When I came back 10 hours later, it was still checking.  After 2 hours
>> more (so it took 12 hours total) it finally finished.  The output was
>
> Anything about erros in dmesg or syslog?

Nope.  All clean.


>> [...]
>> Were there any bad blocks or not?  Is
>> there a way to query the filesystem for the now known bad blocks?  (The
>> "Updating bad block inode." message suggests that such a list is stored
>> directly inside the filesystem.)
>
> dumpe2fs -b /dev/sdb1 probably also works for ext4.

Thanks.  I just tried and it prints nothing.  I guess that means no bad 
blocks were found.

(Rant: Don't you just love programs that instead of explicitly telling 
you that all is OK, they just stay silent, leaving you wondering whether 
they actually work at all?  Argh...)

Re: [gentoo-user] How do I show list of bad blocks on a disk?

[Florian Philipp]

[-- Attachment #1: Type: text/plain, Size: 1322 bytes --]

Am 06.03.2011 18:07, schrieb Nikos Chantziaras:
> Before leaving home, I started an fsck.ext4 on a filesystem (500GB) that
> resides on a disk that I suspect is damaged:
> 
>   fsck.ext4 -c -c -f /dev/sdb1
> 
> When I came back 10 hours later, it was still checking.  After 2 hours
> more (so it took 12 hours total) it finally finished.  The output was:
> 
>   e2fsck 1.41.14 (22-Dec-2010)
>   Checking for bad blocks (non-destructive read-write test)
>   Testing with random pattern: done
>   Extra: Updating bad block inode.
>   Pass 1: Checking inodes, blocks, and sizes
>   Pass 2: Checking directory structure
>   Pass 3: Checking directory connectivity
>   Pass 4: Checking reference counts
>   Pass 5: Checking group summary information
> 
>   Extra: ***** FILE SYSTEM WAS MODIFIED *****
>   Extra: 11/30531584 files (0.0% non-contiguous),
>   1966902/122096638 blocks
> 
> I'm not sure how to read this.  Were there any bad blocks or not?  Is
> there a way to query the filesystem for the now known bad blocks?  (The
> "Updating bad block inode." message suggests that such a list is stored
> directly inside the filesystem.)
> 
> 

When there is nothing else reported, there was no error. "FILE SYSTEM
WAS MODIFIED" usually just means that a directory "lost+found" was created.


[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 262 bytes --]

[gentoo-user] Re: How do I show list of bad blocks on a disk?

[Francesco Talamona]

On Sunday 06 March 2011, Florian Philipp wrote:
> Am 06.03.2011 18:07, schrieb Nikos Chantziaras:
> > Before leaving home, I started an fsck.ext4 on a filesystem (500GB)
> > that
> > 
> > resides on a disk that I suspect is damaged:
> >   fsck.ext4 -c -c -f /dev/sdb1
> > 
> > When I came back 10 hours later, it was still checking.  After 2
> > hours
> > 
> > more (so it took 12 hours total) it finally finished.  The output was:
> >   e2fsck 1.41.14 (22-Dec-2010)
> >   Checking for bad blocks (non-destructive read-write test)
> >   Testing with random pattern: done
> >   Extra: Updating bad block inode.
> >   Pass 1: Checking inodes, blocks, and sizes
> >   Pass 2: Checking directory structure
> >   Pass 3: Checking directory connectivity
> >   Pass 4: Checking reference counts
> >   Pass 5: Checking group summary information
> >   
> >   Extra: ***** FILE SYSTEM WAS MODIFIED *****
> >   Extra: 11/30531584 files (0.0% non-contiguous),
> >   1966902/122096638 blocks
> > 
> > I'm not sure how to read this.  Were there any bad blocks or not? 
> > Is there a way to query the filesystem for the now known bad
> > blocks?  (The "Updating bad block inode." message suggests that
> > such a list is stored directly inside the filesystem.)
> 
> When there is nothing else reported, there was no error. "FILE SYSTEM
> WAS MODIFIED" usually just means that a directory "lost+found" was
> created.

That would be interactive, and it would show up in the console output:

fsck from util-linux-ng 2.18
e2fsck 1.41.12 (17-May-2010)
Pass 1: Checking inodes, blocks, and sizes
Pass 2: Checking directory structure
Pass 3: Checking directory connectivity
/lost+found not found.  Create<y>? yes

Pass 3A: Optimizing directories
Pass 4: Checking reference counts
Pass 5: Checking group summary information

/dev/mapper/sda5: ***** FILE SYSTEM WAS MODIFIED *****
/dev/mapper/sda5: 177646/4481024 files (6.7% non-contiguous), 
10916521/17920370 blocks

Anyway I don't worry about the fact that the filesystem was modified, as 
long as the program doesn't ask for user intervention. As you can see in 
my case there was a directory optimization.

Fsck took a very long time because of "-c" option (you are not taking 
advantage of the fact that the disk is almost empty), and you specified 
it twice, so "the bad block scan will be done using a non-destructive 
read-write test." as stated in the man page, so in the end, nothing to 
worry about WRT filesystem. 

You should also check SMART status.

Bye
	Francesco

-- 
Linux Version 2.6.37-gentoo-r1, Compiled #4 SMP PREEMPT Sat Mar 5 
16:45:57 CET 2011
Two 2.8GHz AMD Athlon 64 Processors, 4GB RAM, 11255 Bogomips Total
aemaeth

[gentoo-user] Re: How do I show list of bad blocks on a disk?

[james]

Nikos Chantziaras <realnc <at> arcor.de> writes:


> Before leaving home, I started an fsck.ext4 on a filesystem (500GB) that 
> resides on a disk that I suspect is damaged:

smartmontools may be of interest to you....

http://smartmontools.sourceforge.net/badblockhowto.html


hth,
James

Re: [gentoo-user] How do I show list of bad blocks on a disk?

[Thanasis]

on 03/06/2011 07:07 PM Nikos Chantziaras wrote the following:
> Before leaving home, I started an fsck.ext4 on a filesystem (500GB)
> that resides on a disk that I suspect is damaged:
>
> fsck.ext4 -c -c -f /dev/sdb1
>
You can check the SMART (Self-Monitoring, Analysis, and Reporting
Technology) data, which is built-in in all hard drives today, and allows
you to see the status or overall “health” of a hard drive.

( smartctl -a /dev/sdb )

Re: [gentoo-user] How do I show list of bad blocks on a disk?

[Dale]

Thanasis wrote:
> on 03/06/2011 07:07 PM Nikos Chantziaras wrote the following:
>    
>> Before leaving home, I started an fsck.ext4 on a filesystem (500GB)
>> that resides on a disk that I suspect is damaged:
>>
>> fsck.ext4 -c -c -f /dev/sdb1
>>
>>      
> You can check the SMART (Self-Monitoring, Analysis, and Reporting
> Technology) data, which is built-in in all hard drives today, and allows
> you to see the status or overall “health” of a hard drive.
>
> ( smartctl -a /dev/sdb )
>
>    

You can also tell it to run a selftest too.  Example commands to look into:

smartctl -t long /dev/<your drive here>

and for the results, usually a good while later:

smartctl -l selftest /dev/<your drive here>

If you think heat may be a issue, try this:

smartctl -a /dev/<your drive here> | grep Temp

With all this, you should be able to find out if you are good to go or 
got a drive going bad.

Dale

:-)  :-)