Fwd: strange e2fsck magic number behaviour

Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

I'm currently trying to recover an ext4 filesystem. Last night, during
a resize operation, the system (Ubuntu 12.04 LTS on my fix-stuff usb
stick) locked up hard and eventually crashed. Restarting,
unsurprisingly, gparted offered to check the volume. e2fsck, called
from within gparted, replayed the journal overnight and completed the
resize.

however, where I was expecting a volume with about 3.5GB of free
space, there was now a volume with 32GB free space, a bit more than
50% utilised. inevitably, trying to boot the linux that lives in there
dropped into grub rescue.

going back, I tried to e2fsck it. this reported large numbers of inode
issues and eventually reported clean. I could mount the volume, but
file metadata looked generally broken (lots of ?s). testdisk showed
the partitions were intact, although it claimed the drive was the
wrong size (incorrectly), and found lots of deleted files within my
ecryptfs home folder. It also found the backup superblocks for the
damaged volume.

the first couple I tried were corrupt, but the third was valid. e2fsck
-b [superblock] -y reports fixing a lot of inode things, checksums,
and then restarts.  it then starts to report hunormous numbers of
multiply-claimed blocks.

and now comes the interesting bit - at some point, block 16777215
starts to appear more and more often in the inodes, often duplicated,
until it starts to print out the number 16777215 in a fast loop. in
fact, it looks like it hits some inode and keeps printing block
16777215 to the same very long line (it's generated 500MB of log)

I removed the first inode containing this block via debugfs, without
this helping.

It sticks out that 16777215 is a magic number (the maximum in a 48 bit
address space) and I google that either ext4 or e2fsck has had a bug
involving it before.

Re: strange e2fsck magic number behaviour

[Alexander Harrowell]

To be clearer, I meant 24 bits.

On Thu, Sep 12, 2013 at 4:39 PM, Alexander Harrowell
<a.harrowell@gmail.com> wrote:
> I'm currently trying to recover an ext4 filesystem. Last night, during
> a resize operation, the system (Ubuntu 12.04 LTS on my fix-stuff usb
> stick) locked up hard and eventually crashed. Restarting,
> unsurprisingly, gparted offered to check the volume. e2fsck, called
> from within gparted, replayed the journal overnight and completed the
> resize.
>
> however, where I was expecting a volume with about 3.5GB of free
> space, there was now a volume with 32GB free space, a bit more than
> 50% utilised. inevitably, trying to boot the linux that lives in there
> dropped into grub rescue.
>
> going back, I tried to e2fsck it. this reported large numbers of inode
> issues and eventually reported clean. I could mount the volume, but
> file metadata looked generally broken (lots of ?s). testdisk showed
> the partitions were intact, although it claimed the drive was the
> wrong size (incorrectly), and found lots of deleted files within my
> ecryptfs home folder. It also found the backup superblocks for the
> damaged volume.
>
> the first couple I tried were corrupt, but the third was valid. e2fsck
> -b [superblock] -y reports fixing a lot of inode things, checksums,
> and then restarts.  it then starts to report hunormous numbers of
> multiply-claimed blocks.
>
> and now comes the interesting bit - at some point, block 16777215
> starts to appear more and more often in the inodes, often duplicated,
> until it starts to print out the number 16777215 in a fast loop. in
> fact, it looks like it hits some inode and keeps printing block
> 16777215 to the same very long line (it's generated 500MB of log)
>
> I removed the first inode containing this block via debugfs, without
> this helping.
>
> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
> address space) and I google that either ext4 or e2fsck has had a bug
> involving it before.

Re: Fwd: strange e2fsck magic number behaviour

[Eric Sandeen]

On 9/12/13 11:39 AM, Alexander Harrowell wrote:
> I'm currently trying to recover an ext4 filesystem. Last night, during
> a resize operation,

from what size to what size? On what kernel?

> the system (Ubuntu 12.04 LTS on my fix-stuff usb
> stick) locked up hard and eventually crashed. Restarting,
> unsurprisingly, gparted offered to check the volume. e2fsck, called
> from within gparted, replayed the journal overnight and completed the
> resize.

hmmm... perhaps.

> however, where I was expecting a volume with about 3.5GB of free
> space, there was now a volume with 32GB free space, a bit more than
> 50% utilised. inevitably, trying to boot the linux that lives in there
> dropped into grub rescue.
> 
> going back, I tried to e2fsck it. this reported large numbers of inode
> issues and eventually reported clean. I could mount the volume, but
> file metadata looked generally broken (lots of ?s). testdisk showed
> the partitions were intact, although it claimed the drive was the
> wrong size (incorrectly), and found lots of deleted files within my
> ecryptfs home folder. It also found the backup superblocks for the
> damaged volume.
> 
> the first couple I tried were corrupt, but the third was valid. e2fsck
> -b [superblock] -y reports fixing a lot of inode things, checksums,
> and then restarts.  it then starts to report hunormous numbers of
> multiply-claimed blocks.
> 
> and now comes the interesting bit - at some point, block 16777215
> starts to appear more and more often in the inodes, often duplicated,
> until it starts to print out the number 16777215 in a fast loop. in
> fact, it looks like it hits some inode and keeps printing block
> 16777215 to the same very long line (it's generated 500MB of log)

= 111111111111111111111111 binary.

Guessing it's maybe a bitmap block?

Resize2fs has had a lot of trouble lately it seems.  You may have just
been the unlucky recipient of a resize2fs bug...

-Eric

> I removed the first inode containing this block via debugfs, without
> this helping.
> 
> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
> address space) and I google that either ext4 or e2fsck has had a bug
> involving it before.
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Fwd: Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

---------- Forwarded message ----------
From: Alexander Harrowell <a.harrowell@gmail.com>
Date: Thu, Sep 12, 2013 at 4:54 PM
Subject: Re: Fwd: strange e2fsck magic number behaviour
To: Eric Sandeen <sandeen@redhat.com>


It was 63GB and I just wanted to fork over 3GB of extra space from my
Windows partition...

The fstab is as follows

/dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
/dev/sda2 Windows7_OS ntfs 63.4G
/dev/sda4 extended partition containing:
-- /dev/sda6 swap linux-swap 8.05G
-- /dev/sda5 /home ext4 66.14G
/dev/sda3 Lenovo_Recovery ntfs 10.25G
unallocated 1M

that's what was intended and is what gparted reports. (however,
weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
/dev/sda4 is correspondingly bigger. this I have only just noticed.)

kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.

thanks for your help.


On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>> I'm currently trying to recover an ext4 filesystem. Last night, during
>> a resize operation,
>
> from what size to what size? On what kernel?
>
>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>> stick) locked up hard and eventually crashed. Restarting,
>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>> from within gparted, replayed the journal overnight and completed the
>> resize.
>
> hmmm... perhaps.
>
>> however, where I was expecting a volume with about 3.5GB of free
>> space, there was now a volume with 32GB free space, a bit more than
>> 50% utilised. inevitably, trying to boot the linux that lives in there
>> dropped into grub rescue.
>>
>> going back, I tried to e2fsck it. this reported large numbers of inode
>> issues and eventually reported clean. I could mount the volume, but
>> file metadata looked generally broken (lots of ?s). testdisk showed
>> the partitions were intact, although it claimed the drive was the
>> wrong size (incorrectly), and found lots of deleted files within my
>> ecryptfs home folder. It also found the backup superblocks for the
>> damaged volume.
>>
>> the first couple I tried were corrupt, but the third was valid. e2fsck
>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>> and then restarts.  it then starts to report hunormous numbers of
>> multiply-claimed blocks.
>>
>> and now comes the interesting bit - at some point, block 16777215
>> starts to appear more and more often in the inodes, often duplicated,
>> until it starts to print out the number 16777215 in a fast loop. in
>> fact, it looks like it hits some inode and keeps printing block
>> 16777215 to the same very long line (it's generated 500MB of log)
>
> = 111111111111111111111111 binary.
>
> Guessing it's maybe a bitmap block?
>
> Resize2fs has had a lot of trouble lately it seems.  You may have just
> been the unlucky recipient of a resize2fs bug...
>
> -Eric
>
>> I removed the first inode containing this block via debugfs, without
>> this helping.
>>
>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>> address space) and I google that either ext4 or e2fsck has had a bug
>> involving it before.
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Eric Sandeen]

On 9/12/13 11:56 AM, Alexander Harrowell wrote:
> ---------- Forwarded message ----------
> From: Alexander Harrowell <a.harrowell@gmail.com>
> Date: Thu, Sep 12, 2013 at 4:54 PM
> Subject: Re: Fwd: strange e2fsck magic number behaviour
> To: Eric Sandeen <sandeen@redhat.com>
> 
> 
> It was 63GB and I just wanted to fork over 3GB of extra space from my
> Windows partition...

Ok, so you tried to resize from 63G to 66G?  Should have been relatively
easy/safe.  I forgot to ask which version of e2fsprogs you had, but if
you did the grow online/mounted, most of the work is done in the kernel.

As Ted said, knowing more info might yield clues:

1) what e2fsprogs version?
2) what were the kernel messages when it crashed/hung?
3) what was the fsck output?

If you didn't save that stuff, it makes it harder to do a post-mortem...

> The fstab is as follows
> 
> /dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
> /dev/sda2 Windows7_OS ntfs 63.4G
> /dev/sda4 extended partition containing:
> -- /dev/sda6 swap linux-swap 8.05G
> -- /dev/sda5 /home ext4 66.14G
> /dev/sda3 Lenovo_Recovery ntfs 10.25G
> unallocated 1M
> 
> that's what was intended and is what gparted reports. (however,
> weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
> /dev/sda4 is correspondingly bigger. this I have only just noticed.)

TBH, I have no idea what Ubuntu Disk Utility does.  I'd trust fdisk -lu
output or /proc/partitions for accurate size info.

Oh; 61.14GiB (powers of 2) == 71 GB (powers of 10)

(61.14*1024*1024*1024/1000/1000/1000 = 71)

So Ubuntu Disk Utility is in cahoots w/ the drive manufacturers, and
using more favorable units.  ;)

-Eric

> kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.
> 
> thanks for your help.
> 
> 
> On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>>> I'm currently trying to recover an ext4 filesystem. Last night, during
>>> a resize operation,
>>
>> from what size to what size? On what kernel?
>>
>>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>>> stick) locked up hard and eventually crashed. Restarting,
>>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>>> from within gparted, replayed the journal overnight and completed the
>>> resize.
>>
>> hmmm... perhaps.
>>
>>> however, where I was expecting a volume with about 3.5GB of free
>>> space, there was now a volume with 32GB free space, a bit more than
>>> 50% utilised. inevitably, trying to boot the linux that lives in there
>>> dropped into grub rescue.
>>>
>>> going back, I tried to e2fsck it. this reported large numbers of inode
>>> issues and eventually reported clean. I could mount the volume, but
>>> file metadata looked generally broken (lots of ?s). testdisk showed
>>> the partitions were intact, although it claimed the drive was the
>>> wrong size (incorrectly), and found lots of deleted files within my
>>> ecryptfs home folder. It also found the backup superblocks for the
>>> damaged volume.
>>>
>>> the first couple I tried were corrupt, but the third was valid. e2fsck
>>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>>> and then restarts.  it then starts to report hunormous numbers of
>>> multiply-claimed blocks.
>>>
>>> and now comes the interesting bit - at some point, block 16777215
>>> starts to appear more and more often in the inodes, often duplicated,
>>> until it starts to print out the number 16777215 in a fast loop. in
>>> fact, it looks like it hits some inode and keeps printing block
>>> 16777215 to the same very long line (it's generated 500MB of log)
>>
>> = 111111111111111111111111 binary.
>>
>> Guessing it's maybe a bitmap block?
>>
>> Resize2fs has had a lot of trouble lately it seems.  You may have just
>> been the unlucky recipient of a resize2fs bug...
>>
>> -Eric
>>
>>> I removed the first inode containing this block via debugfs, without
>>> this helping.
>>>
>>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>>> address space) and I google that either ext4 or e2fsck has had a bug
>>> involving it before.
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html
> 

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

investigating dmesg, I think e2fsck may have been running out of memory.

On Thu, Sep 12, 2013 at 6:59 PM, Eric Sandeen <sandeen@redhat.com> wrote:
> On 9/12/13 11:56 AM, Alexander Harrowell wrote:
>> ---------- Forwarded message ----------
>> From: Alexander Harrowell <a.harrowell@gmail.com>
>> Date: Thu, Sep 12, 2013 at 4:54 PM
>> Subject: Re: Fwd: strange e2fsck magic number behaviour
>> To: Eric Sandeen <sandeen@redhat.com>
>>
>>
>> It was 63GB and I just wanted to fork over 3GB of extra space from my
>> Windows partition...
>
> Ok, so you tried to resize from 63G to 66G?  Should have been relatively
> easy/safe.  I forgot to ask which version of e2fsprogs you had, but if
> you did the grow online/mounted, most of the work is done in the kernel.
>
> As Ted said, knowing more info might yield clues:
>
> 1) what e2fsprogs version?
> 2) what were the kernel messages when it crashed/hung?
> 3) what was the fsck output?
>
> If you didn't save that stuff, it makes it harder to do a post-mortem...
>
>> The fstab is as follows
>>
>> /dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
>> /dev/sda2 Windows7_OS ntfs 63.4G
>> /dev/sda4 extended partition containing:
>> -- /dev/sda6 swap linux-swap 8.05G
>> -- /dev/sda5 /home ext4 66.14G
>> /dev/sda3 Lenovo_Recovery ntfs 10.25G
>> unallocated 1M
>>
>> that's what was intended and is what gparted reports. (however,
>> weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
>> /dev/sda4 is correspondingly bigger. this I have only just noticed.)
>
> TBH, I have no idea what Ubuntu Disk Utility does.  I'd trust fdisk -lu
> output or /proc/partitions for accurate size info.
>
> Oh; 61.14GiB (powers of 2) == 71 GB (powers of 10)
>
> (61.14*1024*1024*1024/1000/1000/1000 = 71)
>
> So Ubuntu Disk Utility is in cahoots w/ the drive manufacturers, and
> using more favorable units.  ;)
>
> -Eric
>
>> kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.
>>
>> thanks for your help.
>>
>>
>> On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>>>> I'm currently trying to recover an ext4 filesystem. Last night, during
>>>> a resize operation,
>>>
>>> from what size to what size? On what kernel?
>>>
>>>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>>>> stick) locked up hard and eventually crashed. Restarting,
>>>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>>>> from within gparted, replayed the journal overnight and completed the
>>>> resize.
>>>
>>> hmmm... perhaps.
>>>
>>>> however, where I was expecting a volume with about 3.5GB of free
>>>> space, there was now a volume with 32GB free space, a bit more than
>>>> 50% utilised. inevitably, trying to boot the linux that lives in there
>>>> dropped into grub rescue.
>>>>
>>>> going back, I tried to e2fsck it. this reported large numbers of inode
>>>> issues and eventually reported clean. I could mount the volume, but
>>>> file metadata looked generally broken (lots of ?s). testdisk showed
>>>> the partitions were intact, although it claimed the drive was the
>>>> wrong size (incorrectly), and found lots of deleted files within my
>>>> ecryptfs home folder. It also found the backup superblocks for the
>>>> damaged volume.
>>>>
>>>> the first couple I tried were corrupt, but the third was valid. e2fsck
>>>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>>>> and then restarts.  it then starts to report hunormous numbers of
>>>> multiply-claimed blocks.
>>>>
>>>> and now comes the interesting bit - at some point, block 16777215
>>>> starts to appear more and more often in the inodes, often duplicated,
>>>> until it starts to print out the number 16777215 in a fast loop. in
>>>> fact, it looks like it hits some inode and keeps printing block
>>>> 16777215 to the same very long line (it's generated 500MB of log)
>>>
>>> = 111111111111111111111111 binary.
>>>
>>> Guessing it's maybe a bitmap block?
>>>
>>> Resize2fs has had a lot of trouble lately it seems.  You may have just
>>> been the unlucky recipient of a resize2fs bug...
>>>
>>> -Eric
>>>
>>>> I removed the first inode containing this block via debugfs, without
>>>> this helping.
>>>>
>>>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>>>> address space) and I google that either ext4 or e2fsck has had a bug
>>>> involving it before.
>>>> --
>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>> the body of a message to majordomo@vger.kernel.org
>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>
>>>
>> --
>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>> the body of a message to majordomo@vger.kernel.org
>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>
>

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

To update, I've found that a) even with 8GB RAM and 8GB swap, e2fsck
can silently run out of memory.

b) something is clearly wrong in block 16777215.

c) debugfs places that block in inode 409774, in use, with an extent
of 16777212-5 and 10 associated filenames, plus several dozen ext2
directory errors.

d) after a first attempt with the updated (1.42.8) version of
e2fsprogs this morning, the disk is mountable again but not much on it
is accessible and the % usage is still screwy.

e) that said, "new" debugfs and e2fsck seem to find more things to fix.

f) trying to decrypt the filenames, most of them don't get found by
ecryptfs-find but the first one produces a list of the files in /home/
and a lot of find: no such file or directory messages.

g) dumpe2fs -b reports no bad blocks. smart reports drive in good condition.

h) I'm quite tempted to zap 409774.

On Thu, Sep 12, 2013 at 7:33 PM, Alexander Harrowell
<a.harrowell@gmail.com> wrote:
> investigating dmesg, I think e2fsck may have been running out of memory.
>
> On Thu, Sep 12, 2013 at 6:59 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>> On 9/12/13 11:56 AM, Alexander Harrowell wrote:
>>> ---------- Forwarded message ----------
>>> From: Alexander Harrowell <a.harrowell@gmail.com>
>>> Date: Thu, Sep 12, 2013 at 4:54 PM
>>> Subject: Re: Fwd: strange e2fsck magic number behaviour
>>> To: Eric Sandeen <sandeen@redhat.com>
>>>
>>>
>>> It was 63GB and I just wanted to fork over 3GB of extra space from my
>>> Windows partition...
>>
>> Ok, so you tried to resize from 63G to 66G?  Should have been relatively
>> easy/safe.  I forgot to ask which version of e2fsprogs you had, but if
>> you did the grow online/mounted, most of the work is done in the kernel.
>>
>> As Ted said, knowing more info might yield clues:
>>
>> 1) what e2fsprogs version?
>> 2) what were the kernel messages when it crashed/hung?
>> 3) what was the fsck output?
>>
>> If you didn't save that stuff, it makes it harder to do a post-mortem...
>>
>>> The fstab is as follows
>>>
>>> /dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
>>> /dev/sda2 Windows7_OS ntfs 63.4G
>>> /dev/sda4 extended partition containing:
>>> -- /dev/sda6 swap linux-swap 8.05G
>>> -- /dev/sda5 /home ext4 66.14G
>>> /dev/sda3 Lenovo_Recovery ntfs 10.25G
>>> unallocated 1M
>>>
>>> that's what was intended and is what gparted reports. (however,
>>> weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
>>> /dev/sda4 is correspondingly bigger. this I have only just noticed.)
>>
>> TBH, I have no idea what Ubuntu Disk Utility does.  I'd trust fdisk -lu
>> output or /proc/partitions for accurate size info.
>>
>> Oh; 61.14GiB (powers of 2) == 71 GB (powers of 10)
>>
>> (61.14*1024*1024*1024/1000/1000/1000 = 71)
>>
>> So Ubuntu Disk Utility is in cahoots w/ the drive manufacturers, and
>> using more favorable units.  ;)
>>
>> -Eric
>>
>>> kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.
>>>
>>> thanks for your help.
>>>
>>>
>>> On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>>> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>>>>> I'm currently trying to recover an ext4 filesystem. Last night, during
>>>>> a resize operation,
>>>>
>>>> from what size to what size? On what kernel?
>>>>
>>>>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>>>>> stick) locked up hard and eventually crashed. Restarting,
>>>>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>>>>> from within gparted, replayed the journal overnight and completed the
>>>>> resize.
>>>>
>>>> hmmm... perhaps.
>>>>
>>>>> however, where I was expecting a volume with about 3.5GB of free
>>>>> space, there was now a volume with 32GB free space, a bit more than
>>>>> 50% utilised. inevitably, trying to boot the linux that lives in there
>>>>> dropped into grub rescue.
>>>>>
>>>>> going back, I tried to e2fsck it. this reported large numbers of inode
>>>>> issues and eventually reported clean. I could mount the volume, but
>>>>> file metadata looked generally broken (lots of ?s). testdisk showed
>>>>> the partitions were intact, although it claimed the drive was the
>>>>> wrong size (incorrectly), and found lots of deleted files within my
>>>>> ecryptfs home folder. It also found the backup superblocks for the
>>>>> damaged volume.
>>>>>
>>>>> the first couple I tried were corrupt, but the third was valid. e2fsck
>>>>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>>>>> and then restarts.  it then starts to report hunormous numbers of
>>>>> multiply-claimed blocks.
>>>>>
>>>>> and now comes the interesting bit - at some point, block 16777215
>>>>> starts to appear more and more often in the inodes, often duplicated,
>>>>> until it starts to print out the number 16777215 in a fast loop. in
>>>>> fact, it looks like it hits some inode and keeps printing block
>>>>> 16777215 to the same very long line (it's generated 500MB of log)
>>>>
>>>> = 111111111111111111111111 binary.
>>>>
>>>> Guessing it's maybe a bitmap block?
>>>>
>>>> Resize2fs has had a lot of trouble lately it seems.  You may have just
>>>> been the unlucky recipient of a resize2fs bug...
>>>>
>>>> -Eric
>>>>
>>>>> I removed the first inode containing this block via debugfs, without
>>>>> this helping.
>>>>>
>>>>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>>>>> address space) and I google that either ext4 or e2fsck has had a bug
>>>>> involving it before.
>>>>> --
>>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>>> the body of a message to majordomo@vger.kernel.org
>>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>>
>>>>
>>> --
>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>> the body of a message to majordomo@vger.kernel.org
>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>
>>

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

Hmm, coming back to this, block 16777215 with identical content is
recurring at intervals of 8 inodes.

On Fri, Sep 13, 2013 at 11:46 AM, Alexander Harrowell
<a.harrowell@gmail.com> wrote:
> To update, I've found that a) even with 8GB RAM and 8GB swap, e2fsck
> can silently run out of memory.
>
> b) something is clearly wrong in block 16777215.
>
> c) debugfs places that block in inode 409774, in use, with an extent
> of 16777212-5 and 10 associated filenames, plus several dozen ext2
> directory errors.
>
> d) after a first attempt with the updated (1.42.8) version of
> e2fsprogs this morning, the disk is mountable again but not much on it
> is accessible and the % usage is still screwy.
>
> e) that said, "new" debugfs and e2fsck seem to find more things to fix.
>
> f) trying to decrypt the filenames, most of them don't get found by
> ecryptfs-find but the first one produces a list of the files in /home/
> and a lot of find: no such file or directory messages.
>
> g) dumpe2fs -b reports no bad blocks. smart reports drive in good condition.
>
> h) I'm quite tempted to zap 409774.
>
> On Thu, Sep 12, 2013 at 7:33 PM, Alexander Harrowell
> <a.harrowell@gmail.com> wrote:
>> investigating dmesg, I think e2fsck may have been running out of memory.
>>
>> On Thu, Sep 12, 2013 at 6:59 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>> On 9/12/13 11:56 AM, Alexander Harrowell wrote:
>>>> ---------- Forwarded message ----------
>>>> From: Alexander Harrowell <a.harrowell@gmail.com>
>>>> Date: Thu, Sep 12, 2013 at 4:54 PM
>>>> Subject: Re: Fwd: strange e2fsck magic number behaviour
>>>> To: Eric Sandeen <sandeen@redhat.com>
>>>>
>>>>
>>>> It was 63GB and I just wanted to fork over 3GB of extra space from my
>>>> Windows partition...
>>>
>>> Ok, so you tried to resize from 63G to 66G?  Should have been relatively
>>> easy/safe.  I forgot to ask which version of e2fsprogs you had, but if
>>> you did the grow online/mounted, most of the work is done in the kernel.
>>>
>>> As Ted said, knowing more info might yield clues:
>>>
>>> 1) what e2fsprogs version?
>>> 2) what were the kernel messages when it crashed/hung?
>>> 3) what was the fsck output?
>>>
>>> If you didn't save that stuff, it makes it harder to do a post-mortem...
>>>
>>>> The fstab is as follows
>>>>
>>>> /dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
>>>> /dev/sda2 Windows7_OS ntfs 63.4G
>>>> /dev/sda4 extended partition containing:
>>>> -- /dev/sda6 swap linux-swap 8.05G
>>>> -- /dev/sda5 /home ext4 66.14G
>>>> /dev/sda3 Lenovo_Recovery ntfs 10.25G
>>>> unallocated 1M
>>>>
>>>> that's what was intended and is what gparted reports. (however,
>>>> weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
>>>> /dev/sda4 is correspondingly bigger. this I have only just noticed.)
>>>
>>> TBH, I have no idea what Ubuntu Disk Utility does.  I'd trust fdisk -lu
>>> output or /proc/partitions for accurate size info.
>>>
>>> Oh; 61.14GiB (powers of 2) == 71 GB (powers of 10)
>>>
>>> (61.14*1024*1024*1024/1000/1000/1000 = 71)
>>>
>>> So Ubuntu Disk Utility is in cahoots w/ the drive manufacturers, and
>>> using more favorable units.  ;)
>>>
>>> -Eric
>>>
>>>> kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.
>>>>
>>>> thanks for your help.
>>>>
>>>>
>>>> On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>>>> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>>>>>> I'm currently trying to recover an ext4 filesystem. Last night, during
>>>>>> a resize operation,
>>>>>
>>>>> from what size to what size? On what kernel?
>>>>>
>>>>>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>>>>>> stick) locked up hard and eventually crashed. Restarting,
>>>>>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>>>>>> from within gparted, replayed the journal overnight and completed the
>>>>>> resize.
>>>>>
>>>>> hmmm... perhaps.
>>>>>
>>>>>> however, where I was expecting a volume with about 3.5GB of free
>>>>>> space, there was now a volume with 32GB free space, a bit more than
>>>>>> 50% utilised. inevitably, trying to boot the linux that lives in there
>>>>>> dropped into grub rescue.
>>>>>>
>>>>>> going back, I tried to e2fsck it. this reported large numbers of inode
>>>>>> issues and eventually reported clean. I could mount the volume, but
>>>>>> file metadata looked generally broken (lots of ?s). testdisk showed
>>>>>> the partitions were intact, although it claimed the drive was the
>>>>>> wrong size (incorrectly), and found lots of deleted files within my
>>>>>> ecryptfs home folder. It also found the backup superblocks for the
>>>>>> damaged volume.
>>>>>>
>>>>>> the first couple I tried were corrupt, but the third was valid. e2fsck
>>>>>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>>>>>> and then restarts.  it then starts to report hunormous numbers of
>>>>>> multiply-claimed blocks.
>>>>>>
>>>>>> and now comes the interesting bit - at some point, block 16777215
>>>>>> starts to appear more and more often in the inodes, often duplicated,
>>>>>> until it starts to print out the number 16777215 in a fast loop. in
>>>>>> fact, it looks like it hits some inode and keeps printing block
>>>>>> 16777215 to the same very long line (it's generated 500MB of log)
>>>>>
>>>>> = 111111111111111111111111 binary.
>>>>>
>>>>> Guessing it's maybe a bitmap block?
>>>>>
>>>>> Resize2fs has had a lot of trouble lately it seems.  You may have just
>>>>> been the unlucky recipient of a resize2fs bug...
>>>>>
>>>>> -Eric
>>>>>
>>>>>> I removed the first inode containing this block via debugfs, without
>>>>>> this helping.
>>>>>>
>>>>>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>>>>>> address space) and I google that either ext4 or e2fsck has had a bug
>>>>>> involving it before.
>>>>>> --
>>>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>>>> the body of a message to majordomo@vger.kernel.org
>>>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>>>
>>>>>
>>>> --
>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>> the body of a message to majordomo@vger.kernel.org
>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>
>>>

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Alexander Harrowell]

example:

Block    Inode number
16777215    2937846
debugfs:  clri <2937846>
debugfs:  icheck 16777215
Block    Inode number
16777215    2937854
debugfs:  clri <2937854>
debugfs:  icheck 16777215
Block    Inode number
16777215    2937862
debugfs:  clri <2937862>
debugfs:  icheck 16777215
Block    Inode number
16777215    2937870
debugfs:  clri <2937870>
debugfs:  icheck 16777215
Block    Inode number
16777215    2937878
debugfs:  clri <2937878>
debugfs:  icheck 16777215

debugfs:  block_dump 16777215
0000  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
2720  0000 0000 0000 0000 0000 0000 ffff ff00  ................
2740  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
3620  0000 0000 0000 0000 0000 0000 ffff ff00  ................
3640  ffff ff00 ffff ff00 ffff ff00 ffff ff00  ................
*
4000  ffff ff00 ffff ff00 ffff ff00 0000 0000  ................
4020  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
4400  ffff ff00 ffff ff00 ffff ff00 ffff ff00  ................
*
4720  ffff ff00 ffff ff00 ffff ff00 0000 0000  ................
4740  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
5640  ffff ff00 ffff ff00 ffff ff00 ffff ff00  ................
*
6000  ffff ff00 ffff ff00 0000 0000 0000 0000  ................
6020  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
6400  0000 0000 ffff ff00 ffff ff00 ffff ff00  ................
6420  ffff ff00 ffff ff00 ffff ff00 ffff ff00  ................
*
6720  ffff ff00 ffff ff00 0000 0000 0000 0000  ................
6740  0000 0000 0000 0000 0000 0000 0000 0000  ................
*
7640  0000 0000 0000 0000 ffff ff00 ffff ff00  ................
7660  ffff ff00 ffff ff00 ffff ff00 ffff ff00  ................
*
7720  ffff ff00 0000 0000 0000 0000 0000 0000  ................
7740  0000 0000 0000 0000 0000 0000 0000 0000  ................
*

On Fri, Sep 13, 2013 at 1:33 PM, Alexander Harrowell
<a.harrowell@gmail.com> wrote:
> Hmm, coming back to this, block 16777215 with identical content is
> recurring at intervals of 8 inodes.
>
> On Fri, Sep 13, 2013 at 11:46 AM, Alexander Harrowell
> <a.harrowell@gmail.com> wrote:
>> To update, I've found that a) even with 8GB RAM and 8GB swap, e2fsck
>> can silently run out of memory.
>>
>> b) something is clearly wrong in block 16777215.
>>
>> c) debugfs places that block in inode 409774, in use, with an extent
>> of 16777212-5 and 10 associated filenames, plus several dozen ext2
>> directory errors.
>>
>> d) after a first attempt with the updated (1.42.8) version of
>> e2fsprogs this morning, the disk is mountable again but not much on it
>> is accessible and the % usage is still screwy.
>>
>> e) that said, "new" debugfs and e2fsck seem to find more things to fix.
>>
>> f) trying to decrypt the filenames, most of them don't get found by
>> ecryptfs-find but the first one produces a list of the files in /home/
>> and a lot of find: no such file or directory messages.
>>
>> g) dumpe2fs -b reports no bad blocks. smart reports drive in good condition.
>>
>> h) I'm quite tempted to zap 409774.
>>
>> On Thu, Sep 12, 2013 at 7:33 PM, Alexander Harrowell
>> <a.harrowell@gmail.com> wrote:
>>> investigating dmesg, I think e2fsck may have been running out of memory.
>>>
>>> On Thu, Sep 12, 2013 at 6:59 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>>> On 9/12/13 11:56 AM, Alexander Harrowell wrote:
>>>>> ---------- Forwarded message ----------
>>>>> From: Alexander Harrowell <a.harrowell@gmail.com>
>>>>> Date: Thu, Sep 12, 2013 at 4:54 PM
>>>>> Subject: Re: Fwd: strange e2fsck magic number behaviour
>>>>> To: Eric Sandeen <sandeen@redhat.com>
>>>>>
>>>>>
>>>>> It was 63GB and I just wanted to fork over 3GB of extra space from my
>>>>> Windows partition...
>>>>
>>>> Ok, so you tried to resize from 63G to 66G?  Should have been relatively
>>>> easy/safe.  I forgot to ask which version of e2fsprogs you had, but if
>>>> you did the grow online/mounted, most of the work is done in the kernel.
>>>>
>>>> As Ted said, knowing more info might yield clues:
>>>>
>>>> 1) what e2fsprogs version?
>>>> 2) what were the kernel messages when it crashed/hung?
>>>> 3) what was the fsck output?
>>>>
>>>> If you didn't save that stuff, it makes it harder to do a post-mortem...
>>>>
>>>>> The fstab is as follows
>>>>>
>>>>> /dev/sda1 SYSTEM_DRV ntfs 1.17g (boot)
>>>>> /dev/sda2 Windows7_OS ntfs 63.4G
>>>>> /dev/sda4 extended partition containing:
>>>>> -- /dev/sda6 swap linux-swap 8.05G
>>>>> -- /dev/sda5 /home ext4 66.14G
>>>>> /dev/sda3 Lenovo_Recovery ntfs 10.25G
>>>>> unallocated 1M
>>>>>
>>>>> that's what was intended and is what gparted reports. (however,
>>>>> weirdly, if you ask Ubuntu Disk Utility, it says /dev/sda5 is 71GB and
>>>>> /dev/sda4 is correspondingly bigger. this I have only just noticed.)
>>>>
>>>> TBH, I have no idea what Ubuntu Disk Utility does.  I'd trust fdisk -lu
>>>> output or /proc/partitions for accurate size info.
>>>>
>>>> Oh; 61.14GiB (powers of 2) == 71 GB (powers of 10)
>>>>
>>>> (61.14*1024*1024*1024/1000/1000/1000 = 71)
>>>>
>>>> So Ubuntu Disk Utility is in cahoots w/ the drive manufacturers, and
>>>> using more favorable units.  ;)
>>>>
>>>> -Eric
>>>>
>>>>> kernel is 3.2.0-29-generic, machine is a ThinkPad X200s with 160GB disk.
>>>>>
>>>>> thanks for your help.
>>>>>
>>>>>
>>>>> On Thu, Sep 12, 2013 at 4:44 PM, Eric Sandeen <sandeen@redhat.com> wrote:
>>>>>> On 9/12/13 11:39 AM, Alexander Harrowell wrote:
>>>>>>> I'm currently trying to recover an ext4 filesystem. Last night, during
>>>>>>> a resize operation,
>>>>>>
>>>>>> from what size to what size? On what kernel?
>>>>>>
>>>>>>> the system (Ubuntu 12.04 LTS on my fix-stuff usb
>>>>>>> stick) locked up hard and eventually crashed. Restarting,
>>>>>>> unsurprisingly, gparted offered to check the volume. e2fsck, called
>>>>>>> from within gparted, replayed the journal overnight and completed the
>>>>>>> resize.
>>>>>>
>>>>>> hmmm... perhaps.
>>>>>>
>>>>>>> however, where I was expecting a volume with about 3.5GB of free
>>>>>>> space, there was now a volume with 32GB free space, a bit more than
>>>>>>> 50% utilised. inevitably, trying to boot the linux that lives in there
>>>>>>> dropped into grub rescue.
>>>>>>>
>>>>>>> going back, I tried to e2fsck it. this reported large numbers of inode
>>>>>>> issues and eventually reported clean. I could mount the volume, but
>>>>>>> file metadata looked generally broken (lots of ?s). testdisk showed
>>>>>>> the partitions were intact, although it claimed the drive was the
>>>>>>> wrong size (incorrectly), and found lots of deleted files within my
>>>>>>> ecryptfs home folder. It also found the backup superblocks for the
>>>>>>> damaged volume.
>>>>>>>
>>>>>>> the first couple I tried were corrupt, but the third was valid. e2fsck
>>>>>>> -b [superblock] -y reports fixing a lot of inode things, checksums,
>>>>>>> and then restarts.  it then starts to report hunormous numbers of
>>>>>>> multiply-claimed blocks.
>>>>>>>
>>>>>>> and now comes the interesting bit - at some point, block 16777215
>>>>>>> starts to appear more and more often in the inodes, often duplicated,
>>>>>>> until it starts to print out the number 16777215 in a fast loop. in
>>>>>>> fact, it looks like it hits some inode and keeps printing block
>>>>>>> 16777215 to the same very long line (it's generated 500MB of log)
>>>>>>
>>>>>> = 111111111111111111111111 binary.
>>>>>>
>>>>>> Guessing it's maybe a bitmap block?
>>>>>>
>>>>>> Resize2fs has had a lot of trouble lately it seems.  You may have just
>>>>>> been the unlucky recipient of a resize2fs bug...
>>>>>>
>>>>>> -Eric
>>>>>>
>>>>>>> I removed the first inode containing this block via debugfs, without
>>>>>>> this helping.
>>>>>>>
>>>>>>> It sticks out that 16777215 is a magic number (the maximum in a 48 bit
>>>>>>> address space) and I google that either ext4 or e2fsck has had a bug
>>>>>>> involving it before.
>>>>>>> --
>>>>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>>>>> the body of a message to majordomo@vger.kernel.org
>>>>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>>>>
>>>>>>
>>>>> --
>>>>> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
>>>>> the body of a message to majordomo@vger.kernel.org
>>>>> More majordomo info at  http://vger.kernel.org/majordomo-info.html
>>>>>
>>>>

Re: Fwd: Fwd: strange e2fsck magic number behaviour

[Theodore Ts'o]

On Fri, Sep 13, 2013 at 01:33:12PM +0000, Alexander Harrowell wrote:
> Hmm, coming back to this, block 16777215 with identical content is
> recurring at intervals of 8 inodes.

So you might want to check and see if there are overlapping metadata
blocks --- that is, a bitmap allocation block that is also part of the
inode table block, or multiple block groups that point at the same
place for their inode table block.

The other thing is before you do more experimentation, I hope you have
made an image backup of your disk.  The more you play games by running
clri and then re-running e2fsck, the more likely that you might
accidentally do damage that might cause less data to be recovered.

In general, especially when the file system is this small such that
it's relatively easy to do an image level backup, the moment that you
think something might have gone off the rails, the wisest thing to do
is to make an image-level backup of the partition before you try to
repair things.

The other thing that has to be asked here is how much do you care
about this 64GB worth of data?  How much is OS data that can be easily
reproduced via an install, and how much are things like your home
directory?  And how recent was your last backup?  It may be that it's
not worth doing a whole lot more work trying to figure out what was
going on.

The other thing is that if this file system is this small, would you
be willing to use e2image to send me a copy of the metadata blocks, so
I can take a look at it myself.  No guarantees that I will find
anything useful, but I'll probably get more information that way.

Regards,

							- Ted

Re: Fwd: strange e2fsck magic number behaviour

[Theodore Ts'o]

On Thu, Sep 12, 2013 at 04:39:33PM +0000, Alexander Harrowell wrote:
> I'm currently trying to recover an ext4 filesystem. Last night, during
> a resize operation, the system (Ubuntu 12.04 LTS on my fix-stuff usb
> stick) locked up hard and eventually crashed. Restarting,
> unsurprisingly, gparted offered to check the volume. e2fsck, called
> from within gparted, replayed the journal overnight and completed the
> resize.

How big was this file system?  And it sounds like you were doing an
on-line resize (that is, the file system was mounted at the time when
you did the resize)?  There were some bugs there with file file
systems with block numbers > 32-bits (i.e., greater than 16TB).  But
for smaller file systems, online resize should have been fairly safe.
Certainly I'm not aware of any bugs that resulted in the system
locking up hard.

> and now comes the interesting bit - at some point, block 16777215
> starts to appear more and more often in the inodes, often duplicated,
> until it starts to print out the number 16777215 in a fast loop. in
> fact, it looks like it hits some inode and keeps printing block
> 16777215 to the same very long line (it's generated 500MB of log)

0xFFFFFF or 0x1000000 isn't a magic boundary as far as ext4 is
concerned.  It appears that this is showing up as part of the multiply
claimed blocks error message?  That usually happens because there was
garbage in an indirect block or in the extent tree.

What you might have remembered is that the maximum number of physical
blocks with ext4 is 48 bits, but what you are reporting is 24 bits,
which is something else quite different.

It would help to see a short except of exactly what e2fsck reported,
so we could see whether it is being reported as a logical block number
or a physical block number.  However, I suspect this is really much
more of a symptom rather than the cause.

Regards,

						- Ted