[CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

Looks like I sent the following message to an obsolete mailing list...

---------- Forwarded message ----------
Date: Thu, 3 Mar 2005 22:33:40 -0800 (PST)
From: Junfeng Yang <yjf@stanford.edu>
To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
     ext2-devel@lists.sourceforge.net,
     jfs-discussion@www-124.southbury.usf.ibm.com, reiser@namesys.com
Cc: mc@cs.stanford.edu
Subject: [MC] [CHECKER] Do ext2,
     jfs and reiserfs respect mount -o sync/dirsync option?


Hi,

FiSC (our file system checker) emits several warnings on ext2, jfs and
reiserfs, complaining that diretories or files are lost while FiSC
believes they should already be persistent on disk. (ext3 behaves
correctly.)

All warnings boil down to a single cause:  when these file systems are
mounted -o sync or dirsync, dirty blocks are still written out
asynchronously.  It appears to me that these mount options don't have any
effect on these file systems.  Is this the intended behavior?

man mount shows:

              sync   All  I/O to the file system should be done
synchronously.

              dirsync
                     All directory updates within the file  system  should
be
                     done  synchronously.   This  affects the following
system
                     calls: creat, link, unlink, symlink, mkdir, rmdir,
mknod
                     and rename.

Any clafirication on this would be very helpful,

-Junfeng


_______________________________________________
MC mailing list
MC@keeda.stanford.edu
http://keeda.stanford.edu/cgi-bin/mailman/listinfo/mc

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Vladimir Saveliev]

Hello

On Sat, 2005-03-05 at 02:23, Junfeng Yang wrote:
> Looks like I sent the following message to an obsolete mailing list...
> 
> ---------- Forwarded message ----------
> Date: Thu, 3 Mar 2005 22:33:40 -0800 (PST)
> From: Junfeng Yang <yjf@stanford.edu>
> To: Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
>      ext2-devel@lists.sourceforge.net,
>      jfs-discussion@www-124.southbury.usf.ibm.com, reiser@namesys.com
> Cc: mc@cs.stanford.edu
> Subject: [MC] [CHECKER] Do ext2,
>      jfs and reiserfs respect mount -o sync/dirsync option?
> 
> 
> Hi,
> 
> FiSC (our file system checker) emits several warnings on ext2, jfs and
> reiserfs, complaining that diretories or files are lost while FiSC
> believes they should already be persistent on disk. (ext3 behaves
> correctly.)
> 
> All warnings boil down to a single cause:  when these file systems are
> mounted -o sync or dirsync, dirty blocks are still written out
> asynchronously.  It appears to me that these mount options don't have any
> effect on these file systems.  Is this the intended behavior?
> 

I just made the follong test on reiserfs (2.6.11-rc4-mm1):
mkreiserfs /dev/hda6
mount /dev/hda6 /mnt -o sync
touch /mnt/file
mkdir /mnt/d
echo Hello > /mnt/hello
reboot -f -n

When system booted up again I mounted /dev/hda6 and all created things
are there. 

Please describe what should one do to get the problem.

> man mount shows:
> 
>               sync   All  I/O to the file system should be done
> synchronously.
> 
>               dirsync
>                      All directory updates within the file  system  should
> be
>                      done  synchronously.   This  affects the following
> system
>                      calls: creat, link, unlink, symlink, mkdir, rmdir,
> mknod
>                      and rename.
> 
> Any clafirication on this would be very helpful,
> 
> -Junfeng
> 
> 
> _______________________________________________
> MC mailing list
> MC@keeda.stanford.edu
> http://keeda.stanford.edu/cgi-bin/mailman/listinfo/mc
> 
> 

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

> I just made the follong test on reiserfs (2.6.11-rc4-mm1):
> mkreiserfs /dev/hda6
> mount /dev/hda6 /mnt -o sync
> touch /mnt/file
> mkdir /mnt/d
> echo Hello > /mnt/hello
> reboot -f -n

Here is what I do to reproduce the same problem:

1. mkreiserfs on a partition
2. issue several file system operations
3. "crash" and resart the machine
4. run reiserfsck --fix-fixable --yes to recover
5. mount the recovered partition.

It appears that step 4 is _important_ in reproducing the problem.  If I
just mount the crashed disk, everything appears to be fine.  However,
attempt to recover the crashed image using reiserfsck result in
metadata/data loss.

Details are attached below.  Let me know if you need any more information.

The script I use (run as root)
#!/bin/sh
umount /dev/hda9
/sbin/mkreiserfs -f /dev/hda9
mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
ln -s /mnt/sbd1 /mnt/sbd1/0001
touch /mnt/sbd1/0002
mkdir /mnt/sbd1/0003
reboot -f -n

uname -a shows:
Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux

reiserfsck output is:
reiserfsck 3.6.19 (2003 www.namesys.com)

*************************************************************
** If you are using the latest reiserfsprogs and  it fails **
** please  email bug reports to reiserfs-list@namesys.com, **
** providing  as  much  information  as  possible --  your **
** hardware,  kernel,  patches,  settings,  all reiserfsck **
** messages  (including version),  the reiserfsck logfile, **
** check  the  syslog file  for  any  related information. **
** If you would like advice on using this program, support **
** is available  for $25 at  www.namesys.com/support.html. **
*************************************************************

Will check consistency of the filesystem on /dev/hda9
and will fix what can be fixed without --rebuild-tree
Will put log info to 'stdout'
###########
reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
###########
Replaying journal..
No transactions found
Checking internal tree..finished
Comparing bitmaps..finished
Checking Semantic tree:
finished
No corruptions found
There are on the filesystem:
        Leaves 1
        Internal nodes 0
        Directories 1
        Other files 0
        Data block pointers 0 (0 of them are zero)
        Safe links 0
###########
reiserfsck finished at Sat Mar  5 12:16:15 2005
###########

second mount of the crashed disk shows:
ReiserFS: hda9: found reiserfs format "3.6" with standard journal
ReiserFS: hda9: using ordered data mode
ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
ReiserFS: hda9: checking transaction log (hda9)
ReiserFS: hda9: Using r5 hash to sort names

RE: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

On Sat, 5 Mar 2005, Srikanth Venkataraman wrote:

>    What file system checker are you using?

Hi Srikanth, the checker we use are FiSC, which is developed by ourselves.

-Junfeng

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

Hi Vladimir, are you able to reproduce the problem?

Thanks,
-Junfeng

On Sat, 5 Mar 2005, Junfeng Yang wrote:

> > I just made the follong test on reiserfs (2.6.11-rc4-mm1):
> > mkreiserfs /dev/hda6
> > mount /dev/hda6 /mnt -o sync
> > touch /mnt/file
> > mkdir /mnt/d
> > echo Hello > /mnt/hello
> > reboot -f -n
>
> Here is what I do to reproduce the same problem:
>
> 1. mkreiserfs on a partition
> 2. issue several file system operations
> 3. "crash" and resart the machine
> 4. run reiserfsck --fix-fixable --yes to recover
> 5. mount the recovered partition.
>
> It appears that step 4 is _important_ in reproducing the problem.  If I
> just mount the crashed disk, everything appears to be fine.  However,
> attempt to recover the crashed image using reiserfsck result in
> metadata/data loss.
>
> Details are attached below.  Let me know if you need any more information.
>
> The script I use (run as root)
> #!/bin/sh
> umount /dev/hda9
> /sbin/mkreiserfs -f /dev/hda9
> mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
> ln -s /mnt/sbd1 /mnt/sbd1/0001
> touch /mnt/sbd1/0002
> mkdir /mnt/sbd1/0003
> reboot -f -n
>
> uname -a shows:
> Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
>
> reiserfsck output is:
> reiserfsck 3.6.19 (2003 www.namesys.com)
>
> *************************************************************
> ** If you are using the latest reiserfsprogs and  it fails **
> ** please  email bug reports to reiserfs-list@namesys.com, **
> ** providing  as  much  information  as  possible --  your **
> ** hardware,  kernel,  patches,  settings,  all reiserfsck **
> ** messages  (including version),  the reiserfsck logfile, **
> ** check  the  syslog file  for  any  related information. **
> ** If you would like advice on using this program, support **
> ** is available  for $25 at  www.namesys.com/support.html. **
> *************************************************************
>
> Will check consistency of the filesystem on /dev/hda9
> and will fix what can be fixed without --rebuild-tree
> Will put log info to 'stdout'
> ###########
> reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
> ###########
> Replaying journal..
> No transactions found
> Checking internal tree..finished
> Comparing bitmaps..finished
> Checking Semantic tree:
> finished
> No corruptions found
> There are on the filesystem:
>         Leaves 1
>         Internal nodes 0
>         Directories 1
>         Other files 0
>         Data block pointers 0 (0 of them are zero)
>         Safe links 0
> ###########
> reiserfsck finished at Sat Mar  5 12:16:15 2005
> ###########
>
> second mount of the crashed disk shows:
> ReiserFS: hda9: found reiserfs format "3.6" with standard journal
> ReiserFS: hda9: using ordered data mode
> ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
> ReiserFS: hda9: checking transaction log (hda9)
> ReiserFS: hda9: Using r5 hash to sort names
>
>

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

Let me know if you need any more information to reproduce the warning.  I
would really appreciate it if you can cc me once you figure out if it is a
bug.

-Junfeng

On Sun, 6 Mar 2005, Junfeng Yang wrote:

>
> Hi Vladimir, are you able to reproduce the problem?
>
> Thanks,
> -Junfeng
>
> On Sat, 5 Mar 2005, Junfeng Yang wrote:
>
> > > I just made the follong test on reiserfs (2.6.11-rc4-mm1):
> > > mkreiserfs /dev/hda6
> > > mount /dev/hda6 /mnt -o sync
> > > touch /mnt/file
> > > mkdir /mnt/d
> > > echo Hello > /mnt/hello
> > > reboot -f -n
> >
> > Here is what I do to reproduce the same problem:
> >
> > 1. mkreiserfs on a partition
> > 2. issue several file system operations
> > 3. "crash" and resart the machine
> > 4. run reiserfsck --fix-fixable --yes to recover
> > 5. mount the recovered partition.
> >
> > It appears that step 4 is _important_ in reproducing the problem.  If I
> > just mount the crashed disk, everything appears to be fine.  However,
> > attempt to recover the crashed image using reiserfsck result in
> > metadata/data loss.
> >
> > Details are attached below.  Let me know if you need any more information.
> >
> > The script I use (run as root)
> > #!/bin/sh
> > umount /dev/hda9
> > /sbin/mkreiserfs -f /dev/hda9
> > mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
> > ln -s /mnt/sbd1 /mnt/sbd1/0001
> > touch /mnt/sbd1/0002
> > mkdir /mnt/sbd1/0003
> > reboot -f -n
> >
> > uname -a shows:
> > Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
> >
> > reiserfsck output is:
> > reiserfsck 3.6.19 (2003 www.namesys.com)
> >
> > *************************************************************
> > ** If you are using the latest reiserfsprogs and  it fails **
> > ** please  email bug reports to reiserfs-list@namesys.com, **
> > ** providing  as  much  information  as  possible --  your **
> > ** hardware,  kernel,  patches,  settings,  all reiserfsck **
> > ** messages  (including version),  the reiserfsck logfile, **
> > ** check  the  syslog file  for  any  related information. **
> > ** If you would like advice on using this program, support **
> > ** is available  for $25 at  www.namesys.com/support.html. **
> > *************************************************************
> >
> > Will check consistency of the filesystem on /dev/hda9
> > and will fix what can be fixed without --rebuild-tree
> > Will put log info to 'stdout'
> > ###########
> > reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
> > ###########
> > Replaying journal..
> > No transactions found
> > Checking internal tree..finished
> > Comparing bitmaps..finished
> > Checking Semantic tree:
> > finished
> > No corruptions found
> > There are on the filesystem:
> >         Leaves 1
> >         Internal nodes 0
> >         Directories 1
> >         Other files 0
> >         Data block pointers 0 (0 of them are zero)
> >         Safe links 0
> > ###########
> > reiserfsck finished at Sat Mar  5 12:16:15 2005
> > ###########
> >
> > second mount of the crashed disk shows:
> > ReiserFS: hda9: found reiserfs format "3.6" with standard journal
> > ReiserFS: hda9: using ordered data mode
> > ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
> > ReiserFS: hda9: checking transaction log (hda9)
> > ReiserFS: hda9: Using r5 hash to sort names
> >
> >
>
>

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

Hi, our mail server had some problems the last few days.  I'm not sure if
you guys have received my message or not.  Can you please send me an ACK,
even if you haven't gotten time to diagnose the error yet?

Thanks a lot,
-Junfeng

On Wed, 9 Mar 2005, Junfeng Yang wrote:

>
> Let me know if you need any more information to reproduce the warning.  I
> would really appreciate it if you can cc me once you figure out if it is a
> bug.
>
> -Junfeng
>
> On Sun, 6 Mar 2005, Junfeng Yang wrote:
>
> >
> > Hi Vladimir, are you able to reproduce the problem?
> >
> > Thanks,
> > -Junfeng
> >
> > On Sat, 5 Mar 2005, Junfeng Yang wrote:
> >
> > > > I just made the follong test on reiserfs (2.6.11-rc4-mm1):
> > > > mkreiserfs /dev/hda6
> > > > mount /dev/hda6 /mnt -o sync
> > > > touch /mnt/file
> > > > mkdir /mnt/d
> > > > echo Hello > /mnt/hello
> > > > reboot -f -n
> > >
> > > Here is what I do to reproduce the same problem:
> > >
> > > 1. mkreiserfs on a partition
> > > 2. issue several file system operations
> > > 3. "crash" and resart the machine
> > > 4. run reiserfsck --fix-fixable --yes to recover
> > > 5. mount the recovered partition.
> > >
> > > It appears that step 4 is _important_ in reproducing the problem.  If I
> > > just mount the crashed disk, everything appears to be fine.  However,
> > > attempt to recover the crashed image using reiserfsck result in
> > > metadata/data loss.
> > >
> > > Details are attached below.  Let me know if you need any more information.
> > >
> > > The script I use (run as root)
> > > #!/bin/sh
> > > umount /dev/hda9
> > > /sbin/mkreiserfs -f /dev/hda9
> > > mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
> > > ln -s /mnt/sbd1 /mnt/sbd1/0001
> > > touch /mnt/sbd1/0002
> > > mkdir /mnt/sbd1/0003
> > > reboot -f -n
> > >
> > > uname -a shows:
> > > Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
> > >
> > > reiserfsck output is:
> > > reiserfsck 3.6.19 (2003 www.namesys.com)
> > >
> > > *************************************************************
> > > ** If you are using the latest reiserfsprogs and  it fails **
> > > ** please  email bug reports to reiserfs-list@namesys.com, **
> > > ** providing  as  much  information  as  possible --  your **
> > > ** hardware,  kernel,  patches,  settings,  all reiserfsck **
> > > ** messages  (including version),  the reiserfsck logfile, **
> > > ** check  the  syslog file  for  any  related information. **
> > > ** If you would like advice on using this program, support **
> > > ** is available  for $25 at  www.namesys.com/support.html. **
> > > *************************************************************
> > >
> > > Will check consistency of the filesystem on /dev/hda9
> > > and will fix what can be fixed without --rebuild-tree
> > > Will put log info to 'stdout'
> > > ###########
> > > reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
> > > ###########
> > > Replaying journal..
> > > No transactions found
> > > Checking internal tree..finished
> > > Comparing bitmaps..finished
> > > Checking Semantic tree:
> > > finished
> > > No corruptions found
> > > There are on the filesystem:
> > >         Leaves 1
> > >         Internal nodes 0
> > >         Directories 1
> > >         Other files 0
> > >         Data block pointers 0 (0 of them are zero)
> > >         Safe links 0
> > > ###########
> > > reiserfsck finished at Sat Mar  5 12:16:15 2005
> > > ###########
> > >
> > > second mount of the crashed disk shows:
> > > ReiserFS: hda9: found reiserfs format "3.6" with standard journal
> > > ReiserFS: hda9: using ordered data mode
> > > ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
> > > ReiserFS: hda9: checking transaction log (hda9)
> > > ReiserFS: hda9: Using r5 hash to sort names
> > >
> > >
> >
> >
>
>

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[E.Gryaznova]

Hello, Junfeng.

Thank you for the report and sorry for delay.

Yes, the problem exists somewhere, but this is not reiserfsck 
--fix-fixable problem.
I just modified the set of commands by the following way:

mkreiserfs dev && mount dev /mnt -o sync && touch /mnt/file && mkdir 
/mnt/d && echo Hello >/mnt/hello && reboot -fn
After boot it is reasonable to pack the reiserfs metada and compare 
"mount" and "reiserfsck --fix-fixable" results on _one_ crashed filesystem.
# debugreiserfs -p dev | gzip -c  > metadata.gz
1.
mount the crashed disk and see data

2.
umount
dd if=/dev/zero of=dev
gunzip -c metadata.gz | debugreiserfs -u dev
reiserfsck --fix-fixable dev
mount the recovered  filesystem and see data

I this case the results of "mount" and "fsck --fix-fixable && mount "are 
equal.
Nevetheless, the "sync" problem exists somewhere. Because (yes, you are 
right), sometimes there are no correct data on filesystem after such 
"crash". I run this test several times and see that sometimes data is 
lost after reboot -fn.

We are working on this problem now.

Thanks,
Lena

Junfeng Yang wrote:

>Hi, our mail server had some problems the last few days.  I'm not sure if
>you guys have received my message or not.  Can you please send me an ACK,
>even if you haven't gotten time to diagnose the error yet?
>
>Thanks a lot,
>-Junfeng
>
>On Wed, 9 Mar 2005, Junfeng Yang wrote:
>
>  
>
>>Let me know if you need any more information to reproduce the warning.  I
>>would really appreciate it if you can cc me once you figure out if it is a
>>bug.
>>
>>-Junfeng
>>
>>On Sun, 6 Mar 2005, Junfeng Yang wrote:
>>
>>    
>>
>>>Hi Vladimir, are you able to reproduce the problem?
>>>
>>>Thanks,
>>>-Junfeng
>>>
>>>On Sat, 5 Mar 2005, Junfeng Yang wrote:
>>>
>>>      
>>>
>>>>>I just made the follong test on reiserfs (2.6.11-rc4-mm1):
>>>>>mkreiserfs /dev/hda6
>>>>>mount /dev/hda6 /mnt -o sync
>>>>>touch /mnt/file
>>>>>mkdir /mnt/d
>>>>>echo Hello > /mnt/hello
>>>>>reboot -f -n
>>>>>          
>>>>>
>>>>Here is what I do to reproduce the same problem:
>>>>
>>>>1. mkreiserfs on a partition
>>>>2. issue several file system operations
>>>>3. "crash" and resart the machine
>>>>4. run reiserfsck --fix-fixable --yes to recover
>>>>5. mount the recovered partition.
>>>>
>>>>It appears that step 4 is _important_ in reproducing the problem.  If I
>>>>just mount the crashed disk, everything appears to be fine.  However,
>>>>attempt to recover the crashed image using reiserfsck result in
>>>>metadata/data loss.
>>>>
>>>>Details are attached below.  Let me know if you need any more information.
>>>>
>>>>The script I use (run as root)
>>>>#!/bin/sh
>>>>umount /dev/hda9
>>>>/sbin/mkreiserfs -f /dev/hda9
>>>>mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
>>>>ln -s /mnt/sbd1 /mnt/sbd1/0001
>>>>touch /mnt/sbd1/0002
>>>>mkdir /mnt/sbd1/0003
>>>>reboot -f -n
>>>>
>>>>uname -a shows:
>>>>Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
>>>>
>>>>reiserfsck output is:
>>>>reiserfsck 3.6.19 (2003 www.namesys.com)
>>>>
>>>>*************************************************************
>>>>** If you are using the latest reiserfsprogs and  it fails **
>>>>** please  email bug reports to reiserfs-list@namesys.com, **
>>>>** providing  as  much  information  as  possible --  your **
>>>>** hardware,  kernel,  patches,  settings,  all reiserfsck **
>>>>** messages  (including version),  the reiserfsck logfile, **
>>>>** check  the  syslog file  for  any  related information. **
>>>>** If you would like advice on using this program, support **
>>>>** is available  for $25 at  www.namesys.com/support.html. **
>>>>*************************************************************
>>>>
>>>>Will check consistency of the filesystem on /dev/hda9
>>>>and will fix what can be fixed without --rebuild-tree
>>>>Will put log info to 'stdout'
>>>>###########
>>>>reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
>>>>###########
>>>>Replaying journal..
>>>>No transactions found
>>>>Checking internal tree..finished
>>>>Comparing bitmaps..finished
>>>>Checking Semantic tree:
>>>>finished
>>>>No corruptions found
>>>>There are on the filesystem:
>>>>        Leaves 1
>>>>        Internal nodes 0
>>>>        Directories 1
>>>>        Other files 0
>>>>        Data block pointers 0 (0 of them are zero)
>>>>        Safe links 0
>>>>###########
>>>>reiserfsck finished at Sat Mar  5 12:16:15 2005
>>>>###########
>>>>
>>>>second mount of the crashed disk shows:
>>>>ReiserFS: hda9: found reiserfs format "3.6" with standard journal
>>>>ReiserFS: hda9: using ordered data mode
>>>>ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
>>>>ReiserFS: hda9: checking transaction log (hda9)
>>>>ReiserFS: hda9: Using r5 hash to sort names
>>>>
>>>>
>>>>        
>>>>
>>>      
>>>
>>    
>>
>
>
>
>  
>

I changed the subject line so that vitaly will read it....

[Hans Reiser]

Junfeng Yang wrote:

>Hi, our mail server had some problems the last few days.  I'm not sure if
>you guys have received my message or not.  Can you please send me an ACK,
>even if you haven't gotten time to diagnose the error yet?
>
>Thanks a lot,
>-Junfeng
>
>On Wed, 9 Mar 2005, Junfeng Yang wrote:
>
>  
>
>>Let me know if you need any more information to reproduce the warning.  I
>>would really appreciate it if you can cc me once you figure out if it is a
>>bug.
>>
>>-Junfeng
>>
>>On Sun, 6 Mar 2005, Junfeng Yang wrote:
>>
>>    
>>
>>>Hi Vladimir, are you able to reproduce the problem?
>>>
>>>Thanks,
>>>-Junfeng
>>>
>>>On Sat, 5 Mar 2005, Junfeng Yang wrote:
>>>
>>>      
>>>
>>>>>I just made the follong test on reiserfs (2.6.11-rc4-mm1):
>>>>>mkreiserfs /dev/hda6
>>>>>mount /dev/hda6 /mnt -o sync
>>>>>touch /mnt/file
>>>>>mkdir /mnt/d
>>>>>echo Hello > /mnt/hello
>>>>>reboot -f -n
>>>>>          
>>>>>
>>>>Here is what I do to reproduce the same problem:
>>>>
>>>>1. mkreiserfs on a partition
>>>>2. issue several file system operations
>>>>3. "crash" and resart the machine
>>>>4. run reiserfsck --fix-fixable --yes to recover
>>>>5. mount the recovered partition.
>>>>
>>>>It appears that step 4 is _important_ in reproducing the problem.  If I
>>>>just mount the crashed disk, everything appears to be fine.  However,
>>>>attempt to recover the crashed image using reiserfsck result in
>>>>metadata/data loss.
>>>>
>>>>Details are attached below.  Let me know if you need any more information.
>>>>
>>>>The script I use (run as root)
>>>>#!/bin/sh
>>>>umount /dev/hda9
>>>>/sbin/mkreiserfs -f /dev/hda9
>>>>mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
>>>>ln -s /mnt/sbd1 /mnt/sbd1/0001
>>>>touch /mnt/sbd1/0002
>>>>mkdir /mnt/sbd1/0003
>>>>reboot -f -n
>>>>
>>>>uname -a shows:
>>>>Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
>>>>
>>>>reiserfsck output is:
>>>>reiserfsck 3.6.19 (2003 www.namesys.com)
>>>>
>>>>*************************************************************
>>>>** If you are using the latest reiserfsprogs and  it fails **
>>>>** please  email bug reports to reiserfs-list@namesys.com, **
>>>>** providing  as  much  information  as  possible --  your **
>>>>** hardware,  kernel,  patches,  settings,  all reiserfsck **
>>>>** messages  (including version),  the reiserfsck logfile, **
>>>>** check  the  syslog file  for  any  related information. **
>>>>** If you would like advice on using this program, support **
>>>>** is available  for $25 at  www.namesys.com/support.html. **
>>>>*************************************************************
>>>>
>>>>Will check consistency of the filesystem on /dev/hda9
>>>>and will fix what can be fixed without --rebuild-tree
>>>>Will put log info to 'stdout'
>>>>###########
>>>>reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
>>>>###########
>>>>Replaying journal..
>>>>No transactions found
>>>>Checking internal tree..finished
>>>>Comparing bitmaps..finished
>>>>Checking Semantic tree:
>>>>finished
>>>>No corruptions found
>>>>There are on the filesystem:
>>>>        Leaves 1
>>>>        Internal nodes 0
>>>>        Directories 1
>>>>        Other files 0
>>>>        Data block pointers 0 (0 of them are zero)
>>>>        Safe links 0
>>>>###########
>>>>reiserfsck finished at Sat Mar  5 12:16:15 2005
>>>>###########
>>>>
>>>>second mount of the crashed disk shows:
>>>>ReiserFS: hda9: found reiserfs format "3.6" with standard journal
>>>>ReiserFS: hda9: using ordered data mode
>>>>ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
>>>>ReiserFS: hda9: checking transaction log (hda9)
>>>>ReiserFS: hda9: Using r5 hash to sort names
>>>>
>>>>
>>>>        
>>>>
>>>      
>>>
>>    
>>
>
>
>
>  
>

Re: [CHECKER] Do ext2, jfs and reiserfs respect mount -o sync/dirsync option? (fwd)

[Junfeng Yang]

Thanks a lot guys.  Please keep me posted on this.
-Junfeng

On Fri, 11 Mar 2005, E.Gryaznova wrote:

> Hello, Junfeng.
>
> Thank you for the report and sorry for delay.
>
> Yes, the problem exists somewhere, but this is not reiserfsck
> --fix-fixable problem.
> I just modified the set of commands by the following way:
>
> mkreiserfs dev && mount dev /mnt -o sync && touch /mnt/file && mkdir
> /mnt/d && echo Hello >/mnt/hello && reboot -fn
> After boot it is reasonable to pack the reiserfs metada and compare
> "mount" and "reiserfsck --fix-fixable" results on _one_ crashed filesystem.
> # debugreiserfs -p dev | gzip -c  > metadata.gz
> 1.
> mount the crashed disk and see data
>
> 2.
> umount
> dd if=/dev/zero of=dev
> gunzip -c metadata.gz | debugreiserfs -u dev
> reiserfsck --fix-fixable dev
> mount the recovered  filesystem and see data
>
> I this case the results of "mount" and "fsck --fix-fixable && mount "are
> equal.
> Nevetheless, the "sync" problem exists somewhere. Because (yes, you are
> right), sometimes there are no correct data on filesystem after such
> "crash". I run this test several times and see that sometimes data is
> lost after reboot -fn.
>
> We are working on this problem now.
>
> Thanks,
> Lena
>
> Junfeng Yang wrote:
>
> >Hi, our mail server had some problems the last few days.  I'm not sure if
> >you guys have received my message or not.  Can you please send me an ACK,
> >even if you haven't gotten time to diagnose the error yet?
> >
> >Thanks a lot,
> >-Junfeng
> >
> >On Wed, 9 Mar 2005, Junfeng Yang wrote:
> >
> >
> >
> >>Let me know if you need any more information to reproduce the warning.  I
> >>would really appreciate it if you can cc me once you figure out if it is a
> >>bug.
> >>
> >>-Junfeng
> >>
> >>On Sun, 6 Mar 2005, Junfeng Yang wrote:
> >>
> >>
> >>
> >>>Hi Vladimir, are you able to reproduce the problem?
> >>>
> >>>Thanks,
> >>>-Junfeng
> >>>
> >>>On Sat, 5 Mar 2005, Junfeng Yang wrote:
> >>>
> >>>
> >>>
> >>>>>I just made the follong test on reiserfs (2.6.11-rc4-mm1):
> >>>>>mkreiserfs /dev/hda6
> >>>>>mount /dev/hda6 /mnt -o sync
> >>>>>touch /mnt/file
> >>>>>mkdir /mnt/d
> >>>>>echo Hello > /mnt/hello
> >>>>>reboot -f -n
> >>>>>
> >>>>>
> >>>>Here is what I do to reproduce the same problem:
> >>>>
> >>>>1. mkreiserfs on a partition
> >>>>2. issue several file system operations
> >>>>3. "crash" and resart the machine
> >>>>4. run reiserfsck --fix-fixable --yes to recover
> >>>>5. mount the recovered partition.
> >>>>
> >>>>It appears that step 4 is _important_ in reproducing the problem.  If I
> >>>>just mount the crashed disk, everything appears to be fine.  However,
> >>>>attempt to recover the crashed image using reiserfsck result in
> >>>>metadata/data loss.
> >>>>
> >>>>Details are attached below.  Let me know if you need any more information.
> >>>>
> >>>>The script I use (run as root)
> >>>>#!/bin/sh
> >>>>umount /dev/hda9
> >>>>/sbin/mkreiserfs -f /dev/hda9
> >>>>mount -t reiserfs /dev/hda9 /mnt/sbd1 -o sync,dirsync
> >>>>ln -s /mnt/sbd1 /mnt/sbd1/0001
> >>>>touch /mnt/sbd1/0002
> >>>>mkdir /mnt/sbd1/0003
> >>>>reboot -f -n
> >>>>
> >>>>uname -a shows:
> >>>>Linux notus 2.6.11 #1 Sat Mar 5 04:39:12 PST 2005 i686 GNU/Linux
> >>>>
> >>>>reiserfsck output is:
> >>>>reiserfsck 3.6.19 (2003 www.namesys.com)
> >>>>
> >>>>*************************************************************
> >>>>** If you are using the latest reiserfsprogs and  it fails **
> >>>>** please  email bug reports to reiserfs-list@namesys.com, **
> >>>>** providing  as  much  information  as  possible --  your **
> >>>>** hardware,  kernel,  patches,  settings,  all reiserfsck **
> >>>>** messages  (including version),  the reiserfsck logfile, **
> >>>>** check  the  syslog file  for  any  related information. **
> >>>>** If you would like advice on using this program, support **
> >>>>** is available  for $25 at  www.namesys.com/support.html. **
> >>>>*************************************************************
> >>>>
> >>>>Will check consistency of the filesystem on /dev/hda9
> >>>>and will fix what can be fixed without --rebuild-tree
> >>>>Will put log info to 'stdout'
> >>>>###########
> >>>>reiserfsck --fix-fixable started at Sat Mar  5 12:16:12 2005
> >>>>###########
> >>>>Replaying journal..
> >>>>No transactions found
> >>>>Checking internal tree..finished
> >>>>Comparing bitmaps..finished
> >>>>Checking Semantic tree:
> >>>>finished
> >>>>No corruptions found
> >>>>There are on the filesystem:
> >>>>        Leaves 1
> >>>>        Internal nodes 0
> >>>>        Directories 1
> >>>>        Other files 0
> >>>>        Data block pointers 0 (0 of them are zero)
> >>>>        Safe links 0
> >>>>###########
> >>>>reiserfsck finished at Sat Mar  5 12:16:15 2005
> >>>>###########
> >>>>
> >>>>second mount of the crashed disk shows:
> >>>>ReiserFS: hda9: found reiserfs format "3.6" with standard journal
> >>>>ReiserFS: hda9: using ordered data mode
> >>>>ReiserFS: hda9: journal params: device hda9, size 8192, journal first block 18, max trans len 1024, max batch 900, max commit age 30, max trans age 30
> >>>>ReiserFS: hda9: checking transaction log (hda9)
> >>>>ReiserFS: hda9: Using r5 hash to sort names
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >>>
> >>
> >>
> >
> >
> >
> >
> >
>
>