a hosed reiserfs

a hosed reiserfs

[Bruce Israel]

OK, I've been hosed.  I had a disk crash, and my ReiserFS /home partition
doesn't mount (/dev/hdf8 on my machine).

When I ran reiserfsck on it, I got an message saying it's a hardware
problem, with an I/O error that it can't read a block (output included
below).

I don't really care about the disk (an 80 GB Western Digital), even though
I bought it within the past year, but I really want to get the data off of
it.

Even though reiserfsck says there's a hardware problem on hdf8, I boot just
fine from that disk (using hdf1 and hdf7, my /boot and / partitions, both
ext3), and I can also successfully mount and read hdf9, another ReiserFS
partion.

I installed a new disk that I'm booting off of until I can figure out how
to deal with this and not screw things further.

I tried to follow the instructions in
http://namesys.com/bad-block-handling.html by copying the partition using
"dd_rescue /dev/hdf8 /home/bad-partition" (where /home is on the new disk).
It ran for a while and created a 4.5 GB file (it's a 20 GB partition, about
15GB used) and then hung my system (completely, couldn't switch VCs, no
other shells responding, etc.).  After powercycling I did it again, and got
the EXACT same behaviour, down to the exact same number of characters
copied into the 4.5GB output file.

It also tried giving dd_rescue that byte count as the starting point, but
that also hung things.

What can I do to recover my files?  Is there something else I can use to
dd_rescue the partition prior to running "reiserfsck --rebuild-tree"?
Maybe incrementing the byte count by 1K and writing in 1K of 0's?  Or
another invocation that would get it to read the data?

Or something else that would get the data back?

Also, if anyone has any advice on *things to avoid* that might potentially
lower the probability of data recovery, I'd appreciate it.  I've been very
cautious in what I do so that I don't exacerbate the problem. Thanks.

Bruce

======== reiserfsck output

reiserfsck 3.6.11 (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 read-only check consistency of the filesystem on /dev/hdf8
Will put log info to 'check.log'

Do you want to run this program?[N/Yes] (note need to type Yes if you do):###########
reiserfsck --check started at Fri Jun 11 00:15:18 2004
###########
Replaying journal..
0 transactions replayed

The problem has occurred looks like a hardware problem.
If you have bad blocks, we advise you to get a new hard
drive, because once you get one bad block that the disk
drive internals cannot hide from your sight, the chances
of getting more are generally said to become much higher
(precise statistics are unknown to us), and this disk drive
is probably not expensive enough for you to risk your time
and data on it. If you don't want to follow that advice,
then if you have just a few bad blocks, try writing to the
bad blocks and see if the drive remaps the bad blocks (that
means it takes a block it has in reserve and allocates it
for use for requests of that block number).  If it cannot
remap the block, this could be quite bad, as it may mean
that so many blocks have gone bad that none remain in 
reserve to allocate.

bread: Cannot read the block (4161536): (Input/output error).


******************************************************
* Warning: The dma on your hard drive is turned off. *
* This may really slow down the fsck process.        *
******************************************************

Re: a hosed reiserfs

[Vitaly Fertman]

Hello, 

On Friday 25 June 2004 19:52, Bruce Israel wrote:
> OK, I've been hosed.  I had a disk crash, and my ReiserFS /home partition
> doesn't mount (/dev/hdf8 on my machine).
>
> When I ran reiserfsck on it, I got an message saying it's a hardware
> problem, with an I/O error that it can't read a block (output included
> below).
>
> I don't really care about the disk (an 80 GB Western Digital), even though
> I bought it within the past year, but I really want to get the data off of
> it.
>
> Even though reiserfsck says there's a hardware problem on hdf8, I boot just
> fine from that disk (using hdf1 and hdf7, my /boot and / partitions, both
> ext3), and I can also successfully mount and read hdf9, another ReiserFS
> partion.
>
> I installed a new disk that I'm booting off of until I can figure out how
> to deal with this and not screw things further.
>
> I tried to follow the instructions in
> http://namesys.com/bad-block-handling.html by copying the partition using
> "dd_rescue /dev/hdf8 /home/bad-partition" (where /home is on the new disk).
> It ran for a while and created a 4.5 GB file (it's a 20 GB partition, about
> 15GB used) and then hung my system (completely, couldn't switch VCs, no
> other shells responding, etc.).  After powercycling I did it again, and got
> the EXACT same behaviour, down to the exact same number of characters
> copied into the 4.5GB output file.
>
> It also tried giving dd_rescue that byte count as the starting point, but
> that also hung things.
>
> What can I do to recover my files?  Is there something else I can use to
> dd_rescue the partition prior to running "reiserfsck --rebuild-tree"?
> Maybe incrementing the byte count by 1K and writing in 1K of 0's?  Or
> another invocation that would get it to read the data?
>
> Or something else that would get the data back?
>
> Also, if anyone has any advice on *things to avoid* that might potentially
> lower the probability of data recovery, I'd appreciate it.  I've been very
> cautious in what I do so that I don't exacerbate the problem. Thanks.
>
> Bruce
>
> ======== reiserfsck output
>
> reiserfsck 3.6.11 (2003 www.namesys.com)

It would be better to use a newer version of reiserfsprogs.

> bread: Cannot read the block (4161536): (Input/output error).

what do you see in the syslog? any related message about IO 
error or about an access beyond end of the device? It could also 
be the case that the partition end is beyond the harddrive size. 
check it please.

-- 
Thanks,
Vitaly Fertman

Re: a hosed reiserfs

[Dieter Nützel]

Am Freitag, 25. Juni 2004 17:52 schrieb Bruce Israel:
> OK, I've been hosed.  I had a disk crash, and my ReiserFS /home partition
> doesn't mount (/dev/hdf8 on my machine).
>
> When I ran reiserfsck on it, I got an message saying it's a hardware
> problem, with an I/O error that it can't read a block (output included
> below).
>
> I don't really care about the disk (an 80 GB Western Digital), even though
> I bought it within the past year, but I really want to get the data off of
> it.
>
> Even though reiserfsck says there's a hardware problem on hdf8, I boot just
> fine from that disk (using hdf1 and hdf7, my /boot and / partitions, both
> ext3), and I can also successfully mount and read hdf9, another ReiserFS
> partion.
>
> I installed a new disk that I'm booting off of until I can figure out how
> to deal with this and not screw things further.
>
> I tried to follow the instructions in
> http://namesys.com/bad-block-handling.html by copying the partition using
> "dd_rescue /dev/hdf8 /home/bad-partition" (where /home is on the new disk).
> It ran for a while and created a 4.5 GB file (it's a 20 GB partition, about
> 15GB used) and then hung my system (completely, couldn't switch VCs, no
> other shells responding, etc.).  After powercycling I did it again, and got
> the EXACT same behaviour, down to the exact same number of characters
> copied into the 4.5GB output file.
>
> It also tried giving dd_rescue that byte count as the starting point, but
> that also hung things.
>
> What can I do to recover my files?  Is there something else I can use to
> dd_rescue the partition prior to running "reiserfsck --rebuild-tree"?
> Maybe incrementing the byte count by 1K and writing in 1K of 0's?  Or
> another invocation that would get it to read the data?

You can try dd_rescue in revers mode, too.
So you could get your "complete" partition eventually back.

Used that several times (some tries in normal and then some in reverse mode).
Then a --rebuild-tree run.

Good luck.

-Dieter

Re: a hosed reiserfs

[Hans Reiser]

Vitaly Fertman wrote:

>Hello, 
>
>On Friday 25 June 2004 19:52, Bruce Israel wrote:
>  
>
>>OK, I've been hosed.  I had a disk crash, and my ReiserFS /home partition
>>doesn't mount (/dev/hdf8 on my machine).
>>
>>When I ran reiserfsck on it, I got an message saying it's a hardware
>>problem, with an I/O error that it can't read a block (output included
>>below).
>>
>>I don't really care about the disk (an 80 GB Western Digital), even though
>>I bought it within the past year, but I really want to get the data off of
>>it.
>>
>>Even though reiserfsck says there's a hardware problem on hdf8, I boot just
>>fine from that disk (using hdf1 and hdf7, my /boot and / partitions, both
>>ext3), and I can also successfully mount and read hdf9, another ReiserFS
>>partion.
>>
>>I installed a new disk that I'm booting off of until I can figure out how
>>to deal with this and not screw things further.
>>
>>I tried to follow the instructions in
>>http://namesys.com/bad-block-handling.html by copying the partition using
>>"dd_rescue /dev/hdf8 /home/bad-partition" (where /home is on the new disk).
>>It ran for a while and created a 4.5 GB file (it's a 20 GB partition, about
>>15GB used) and then hung my system (completely, couldn't switch VCs, no
>>other shells responding, etc.).  After powercycling I did it again, and got
>>the EXACT same behaviour, down to the exact same number of characters
>>copied into the 4.5GB output file.
>>
>>It also tried giving dd_rescue that byte count as the starting point, but
>>that also hung things.
>>
>>What can I do to recover my files?  Is there something else I can use to
>>dd_rescue the partition prior to running "reiserfsck --rebuild-tree"?
>>Maybe incrementing the byte count by 1K and writing in 1K of 0's?  Or
>>another invocation that would get it to read the data?
>>
>>Or something else that would get the data back?
>>
>>Also, if anyone has any advice on *things to avoid* that might potentially
>>lower the probability of data recovery, I'd appreciate it.  I've been very
>>cautious in what I do so that I don't exacerbate the problem. Thanks.
>>
>>Bruce
>>
>>======== reiserfsck output
>>
>>reiserfsck 3.6.11 (2003 www.namesys.com)
>>    
>>
>
>It would be better to use a newer version of reiserfsprogs.
>
>  
>
>>bread: Cannot read the block (4161536): (Input/output error).
>>    
>>
>
>what do you see in the syslog? any related message about IO 
>error or about an access beyond end of the device? It could also 
>be the case that the partition end is beyond the harddrive size. 
>check it please.
>
>  
>
Vitaly, hardware errors are only supported by us if they pay the $25...

Re: a hosed reiserfs

[Bruce Israel]

On Fri, Jun 25, 2004 at 09:54:15AM -0700, Hans Reiser wrote:
>
> Vitaly, hardware errors are only supported by us if they pay the $25...

OK, I have no problem paying $25 if you can help me.  What should I do?

Bruce

Re: a hosed reiserfs

[Bruce Israel]

On Fri, Jun 25, 2004 at 08:06:55PM +0400, Vitaly Fertman wrote:
> Hello, 
> 
> On Friday 25 June 2004 19:52, Bruce Israel wrote:
> 
> It would be better to use a newer version of reiserfsprogs.

OK, when I get home I'll download and install the latest.  3.6.17?

> > bread: Cannot read the block (4161536): (Input/output error).
> 
> what do you see in the syslog? any related message about IO 
> error or about an access beyond end of the device? It could also 
> be the case that the partition end is beyond the harddrive size. 
> check it please.
> 
> -- 

No, that's not the case.  /dev/hdf8 has the problems, but the partition
right after it, hdf9, is also a reiserfs partion, and it doesn't have any
problems.  The partion table for the disk is below.

I'll check out the logs, though I think I had looked in them at the time
and found nothing informative.

Bruce


root@exsolent / # fdisk /dev/hdf

The number of cylinders for this disk is set to 9729.
There is nothing wrong with that, but this is larger than 1024,
and could in certain setups cause problems with:
1) software that runs at boot time (e.g., old versions of LILO)
2) booting and partitioning software from other OSs
   (e.g., DOS FDISK, OS/2 FDISK)

Command (m for help): p

Disk /dev/hdf: 80.0 GB, 80026361856 bytes
255 heads, 63 sectors/track, 9729 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

   Device Boot      Start         End      Blocks   Id  System
/dev/hdf1               1          13      104391   83  Linux
/dev/hdf2              14          75      498015   82  Linux swap
/dev/hdf3              76        9729    77545755    5  Extended
/dev/hdf5              76        1292     9775521   83  Linux
/dev/hdf6            1293        2509     9775521   83  Linux
/dev/hdf7            2510        3726     9775521   83  Linux
/dev/hdf8            3727        6159    19543041   83  Linux
/dev/hdf9            6160        9729    28675993+  83  Linux

Re: a hosed reiserfs

[Hans Reiser]

Bruce Israel wrote:

>On Fri, Jun 25, 2004 at 09:54:15AM -0700, Hans Reiser wrote:
>  
>
>>Vitaly, hardware errors are only supported by us if they pay the $25...
>>    
>>
>
>OK, I have no problem paying $25 if you can help me.  What should I do?
>
>Bruce
>
>
>  
>
www.namesys.com/support.html takes credit cards.  Thanks for being our 
customer.

Hans

Re: a hosed reiserfs

[Bruce Israel]

On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
>
> www.namesys.com/support.html takes credit cards.  Thanks for being our 
> customer.

OK, done.

Here's the current situation:

My partition doesn't mount.  reiserfsck says there are bad blocks.  The
output of 'debugreiserfs' is as follows:

    Filesystem state: consistency is not checked after last mounting

    Reiserfs super block in block 16 on 0x2148 of format 3.6 with standard journal
    Count of blocks on the device: 4885760
    Number of bitmaps: 150
    Blocksize: 4096
    Free blocks (count of blocks - used [journal, bitmaps, data, reserved] blocks): 1095151
    Root block: 9426
    Filesystem is NOT cleanly umounted
    Tree height: 4
    Hash function used to sort names: "r5"
    Objectid map size 602, max 972
    Journal parameters:
	    Device [0x0]
	    Magic [0x4edee6b]
	    Size 8193 blocks (including 1 for journal header) (first block 18)
	    Max transaction length 1024 blocks
	    Max batch size 900 blocks
	    Max commit age 30
    Blocks reserved by journal: 0
    Fs state field: 0x0:
    sb_version: 2
    inode generation number: 2033166
    UUID: af13ddfc-8726-4732-bbf5-db17b6c6671d
    LABEL: 
    Set flags in SB:
	    ATTRIBUTES CLEAN

The partition table looks fine (and hasn't changed in the last 6 months or
so). 

I've tried running dd_rescue to copy the partition over; I've run it both
forward and reverse.  Whenever I run it, it copies for a while and then
hangs my system (Linux 2.6.5-gentoo-r1; also hangs a 2.6.3 kernel).

When I say it hangs, I mean that all other shells and virtual consoles are
non-responsive, and the only thing that works is the reset button on the
front panel.

I also tried /sbin/badblocks which hung the sytem as well.

When I run dd_rescue forward, it copies about 4.5GB (out of 20GB total),
before hanging, resulting in the following file:

4511808 -rw-r-----    1 root     4615569408 Jul  2 23:50 hdf8

I've also run it forward and then reverse on the same file.  It seems to
only copy about 20 Meg or so in the reverse direction.  The most recent
effort resulted in:

4520612 -rw-r-----    1 root     20012073984 Jul  4 12:04 hdf8-2

I've tried giving it starting byte counts of the size (or 1K less), but it
still hangs, just a lot quicker.  Turning on dd_rescue's logging doesn't
help; it appears that the system is locked up before anything is either
written or flushed.

This is a very slow process since I have to reboot after each time I run
it. 

What should I do?

Bruce

Re: a hosed reiserfs

[Hans Reiser]

Bruce Israel wrote:

>On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
>  
>
>>www.namesys.com/support.html takes credit cards.  Thanks for being our 
>>customer.
>>    
>>
>
>OK, done.
>
>Here's the current situation:
>
>My partition doesn't mount.  reiserfsck says there are bad blocks.  The
>output of 'debugreiserfs' is as follows:
>
>    Filesystem state: consistency is not checked after last mounting
>
>    Reiserfs super block in block 16 on 0x2148 of format 3.6 with standard journal
>    Count of blocks on the device: 4885760
>    Number of bitmaps: 150
>    Blocksize: 4096
>    Free blocks (count of blocks - used [journal, bitmaps, data, reserved] blocks): 1095151
>    Root block: 9426
>    Filesystem is NOT cleanly umounted
>    Tree height: 4
>    Hash function used to sort names: "r5"
>    Objectid map size 602, max 972
>    Journal parameters:
>	    Device [0x0]
>	    Magic [0x4edee6b]
>	    Size 8193 blocks (including 1 for journal header) (first block 18)
>	    Max transaction length 1024 blocks
>	    Max batch size 900 blocks
>	    Max commit age 30
>    Blocks reserved by journal: 0
>    Fs state field: 0x0:
>    sb_version: 2
>    inode generation number: 2033166
>    UUID: af13ddfc-8726-4732-bbf5-db17b6c6671d
>    LABEL: 
>    Set flags in SB:
>	    ATTRIBUTES CLEAN
>
>The partition table looks fine (and hasn't changed in the last 6 months or
>so). 
>
>I've tried running dd_rescue to copy the partition over; I've run it both
>forward and reverse.  Whenever I run it, it copies for a while and then
>hangs my system (Linux 2.6.5-gentoo-r1; also hangs a 2.6.3 kernel).
>
>When I say it hangs, I mean that all other shells and virtual consoles are
>non-responsive, and the only thing that works is the reset button on the
>front panel.
>
>I also tried /sbin/badblocks which hung the sytem as well.
>
>When I run dd_rescue forward, it copies about 4.5GB (out of 20GB total),
>before hanging, resulting in the following file:
>
>4511808 -rw-r-----    1 root     4615569408 Jul  2 23:50 hdf8
>
>I've also run it forward and then reverse on the same file.  It seems to
>only copy about 20 Meg or so in the reverse direction.  The most recent
>effort resulted in:
>
>4520612 -rw-r-----    1 root     20012073984 Jul  4 12:04 hdf8-2
>
>I've tried giving it starting byte counts of the size (or 1K less), but it
>still hangs, just a lot quicker.  Turning on dd_rescue's logging doesn't
>help; it appears that the system is locked up before anything is either
>written or flushed.
>
>This is a very slow process since I have to reboot after each time I run
>it. 
>
>What should I do?
>
>Bruce
>
>
>  
>
this is in the expertise area of flx and vitaly, so I must let them 
answer it.

RE: a hosed reiserfs

[Burnes, James]

Speaking of which.  Does ReiserFS receive funding from the Linux Fund?
(The MasterCard which sends a percentage of all of your MC transactions
to open source projects?)



jim burnes
security engineer
great-west, denver
 

> -----Original Message-----
> From: Hans Reiser [mailto:reiser@namesys.com]
> Sent: Sunday, July 04, 2004 4:57 PM
> To: Bruce Israel
> Cc: reiserfs-list@namesys.com; Vitaly Fertman; Alexander Lyamin aka
FLX
> Subject: Re: a hosed reiserfs
> 
> Bruce Israel wrote:
> 
> >On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
> >
> >
> >>www.namesys.com/support.html takes credit cards.  Thanks for being
our
> >>customer.
> >>

Re: a hosed reiserfs

[Hans Reiser]

Burnes, James wrote:

>Speaking of which.  Does ReiserFS receive funding from the Linux Fund?
>(The MasterCard which sends a percentage of all of your MC transactions
>to open source projects?)
>  
>
No we don't, is there real money in the Linux Fund?

>
>
>jim burnes
>security engineer
>great-west, denver
> 
>
>  
>
>>-----Original Message-----
>>From: Hans Reiser [mailto:reiser@namesys.com]
>>Sent: Sunday, July 04, 2004 4:57 PM
>>To: Bruce Israel
>>Cc: reiserfs-list@namesys.com; Vitaly Fertman; Alexander Lyamin aka
>>    
>>
>FLX
>  
>
>>Subject: Re: a hosed reiserfs
>>
>>Bruce Israel wrote:
>>
>>    
>>
>>>On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
>>>
>>>
>>>      
>>>
>>>>www.namesys.com/support.html takes credit cards.  Thanks for being
>>>>        
>>>>
>our
>  
>
>>>>customer.
>>>>
>>>>        
>>>>
>
>
>
>  
>

Re: a hosed reiserfs

[Redeeman]

/me smells a repacker/resizer coming along
On Tue, 2004-07-06 at 12:05 -0700, Hans Reiser wrote:
> Burnes, James wrote:
> 
> >Speaking of which.  Does ReiserFS receive funding from the Linux Fund?
> >(The MasterCard which sends a percentage of all of your MC transactions
> >to open source projects?)
> >  
> >
> No we don't, is there real money in the Linux Fund?
> 
> >
> >
> >jim burnes
> >security engineer
> >great-west, denver
> > 
> >
> >  
> >
> >>-----Original Message-----
> >>From: Hans Reiser [mailto:reiser@namesys.com]
> >>Sent: Sunday, July 04, 2004 4:57 PM
> >>To: Bruce Israel
> >>Cc: reiserfs-list@namesys.com; Vitaly Fertman; Alexander Lyamin aka
> >>    
> >>
> >FLX
> >  
> >
> >>Subject: Re: a hosed reiserfs
> >>
> >>Bruce Israel wrote:
> >>
> >>    
> >>
> >>>On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
> >>>
> >>>
> >>>      
> >>>
> >>>>www.namesys.com/support.html takes credit cards.  Thanks for being
> >>>>        
> >>>>
> >our
> >  
> >
> >>>>customer.
> >>>>
> >>>>        
> >>>>
> >
> >
> >
> >  
> >
> 
> 
-- 
Redeeman <redeeman@metanurb.dk>

RE: a hosed reiserfs

[Burnes, James]

For some reason I haven't received this reply, except through the
forward.

I would assume there is some money in there.  They do seem to fund
various opensource activities.

Maybe filling out an app for grant would be good.  Is there an available
non-profit wrapper around Reiser4?  There must be some sort of shell for
the DARPA funding to have happened.

Thx,

jim burnes
security engineer
great-west, denver
 

> -----Original Message-----
> From: Redeeman [mailto:redeeman@metanurb.dk]
> Sent: Tuesday, July 06, 2004 4:23 PM
> To: Reiserfs Mailinglist
> Subject: Re: a hosed reiserfs
> 
> /me smells a repacker/resizer coming along
> On Tue, 2004-07-06 at 12:05 -0700, Hans Reiser wrote:
> > Burnes, James wrote:
> >
> > >Speaking of which.  Does ReiserFS receive funding from the Linux
Fund?
> > >(The MasterCard which sends a percentage of all of your MC
transactions
> > >to open source projects?)
> > >
> > >
> > No we don't, is there real money in the Linux Fund?
> >
> > >
> > >
> > >jim burnes
> > >security engineer
> > >great-west, denver
> > >
> > >
> > >
> > >
> > >>-----Original Message-----
> > >>From: Hans Reiser [mailto:reiser@namesys.com]
> > >>Sent: Sunday, July 04, 2004 4:57 PM
> > >>To: Bruce Israel
> > >>Cc: reiserfs-list@namesys.com; Vitaly Fertman; Alexander Lyamin
aka
> > >>
> > >>
> > >FLX
> > >
> > >
> > >>Subject: Re: a hosed reiserfs
> > >>
> > >>Bruce Israel wrote:
> > >>
> > >>
> > >>
> > >>>On Mon, Jun 28, 2004 at 10:46:05AM -0700, Hans Reiser wrote:
> > >>>
> > >>>
> > >>>
> > >>>
> > >>>>www.namesys.com/support.html takes credit cards.  Thanks for
being
> > >>>>
> > >>>>
> > >our
> > >
> > >
> > >>>>customer.
> > >>>>
> > >>>>
> > >>>>
> > >
> > >
> > >
> > >
> > >
> >
> >
> --
> Redeeman <redeeman@metanurb.dk>

Re: a hosed reiserfs

[Philip Miller]

Bruce Israel wrote:
> [snip]
> I've tried running dd_rescue to copy the partition over; I've run it both
> forward and reverse.  Whenever I run it, it copies for a while and then
> hangs my system (Linux 2.6.5-gentoo-r1; also hangs a 2.6.3 kernel).
> 
> When I say it hangs, I mean that all other shells and virtual consoles are
> non-responsive, and the only thing that works is the reset button on the
> front panel.
> 
> I also tried /sbin/badblocks which hung the sytem as well.
> 
> When I run dd_rescue forward, it copies about 4.5GB (out of 20GB total),
> before hanging, resulting in the following file:
> 
> 4511808 -rw-r-----    1 root     4615569408 Jul  2 23:50 hdf8
> 
> I've also run it forward and then reverse on the same file.  It seems to
> only copy about 20 Meg or so in the reverse direction.  The most recent
> effort resulted in:
> 
> 4520612 -rw-r-----    1 root     20012073984 Jul  4 12:04 hdf8-2
> 
> I've tried giving it starting byte counts of the size (or 1K less), but it
> still hangs, just a lot quicker.  Turning on dd_rescue's logging doesn't
> help; it appears that the system is locked up before anything is either
> written or flushed.
> 
> This is a very slow process since I have to reboot after each time I run
> it. 

Do you, by any chance, run a system based on the nForce2 chipset? On some 
boards, issues with the APIC reliably cause lockups under heavy I/O load, 
such as dd_rescue would create.

If so, try the following:
With kernel version of 2.6.5 or later, try booting with the kernel command 
line argument 'apic_tack=2'.

Philip Miller

Re: a hosed reiserfs

[mjt]

On Tue, Jul 06, 2004 at 04:40:31PM -0600, Burnes, James wrote:
>
>I would assume there is some money in there.  They do seem to fund
>various opensource activities.

As in events? Or really active software development?

Well, Reiser4 has some stuff that's yet todo and some stuff that'd
be nice to have and maybe even a bug or two, so I guess all money
would be welcome...

If the tail-packing bug would be fixed, or whatever remains, it'd
be great if Namesys wrote the online (profiling) repacker/resizer/fsck(?)
program to generate money. 

This would be even better if cryptcompress work would be done at the same 
time, since that's something people look forward to and something that's 
not quite there yet. Cryptcompress might be one reason for people to move
to Reiser4 and thus more buyers for the repacker.

Since feature-wise I don't see Reiser4 that lacking. Maybe the biggest issue
is of course viewports, but since that's coming by DARPA, there's stuff
like copy-on-capture and chdir to a non-executable file and syscalls.
But nothing mission critical or half-broken. Right?

>Maybe filling out an app for grant would be good.  Is there an available
>non-profit wrapper around Reiser4?  There must be some sort of shell for
>the DARPA funding to have happened.

Non-profit wrapper? You mean some non-profit "welfare" organization?
Why can't DARPA fund a company?
Or did/will they donate countless units of money through WorldPay? ;)

-- 
mjt

Re: a hosed reiserfs

[Hans Reiser]

Burnes, James wrote:

>For some reason I haven't received this reply, except through the
>forward.
>
>I would assume there is some money in there.  They do seem to fund
>various opensource activities.
>
>Maybe filling out an app for grant would be good.  Is there an available
>non-profit wrapper around Reiser4?  There must be some sort of shell for
>the DARPA funding to have happened.
>
>Thx,
>
>jim burnes
>security engineer
>great-west, denver
> 
>  
>
Namesys is a for profit corporation.

Vitaly, can you look into it?  Thanks.

Re: a hosed reiserfs

[Hans Reiser]

Markus Törnqvist wrote:

>On Tue, Jul 06, 2004 at 04:40:31PM -0600, Burnes, James wrote:
>  
>
>>I would assume there is some money in there.  They do seem to fund
>>various opensource activities.
>>    
>>
>
>As in events? Or really active software development?
>
>Well, Reiser4 has some stuff that's yet todo and some stuff that'd
>be nice to have and maybe even a bug or two, so I guess all money
>would be welcome...
>
>If the tail-packing bug would be fixed, or whatever remains, it'd
>be great if Namesys wrote the online (profiling) repacker/resizer/fsck(?)
>program to generate money. 
>
>This would be even better if cryptcompress work would be done at the same 
>time, since that's something people look forward to and something that's 
>not quite there yet. Cryptcompress might be one reason for people to move
>to Reiser4 and thus more buyers for the repacker.
>
>Since feature-wise I don't see Reiser4 that lacking. Maybe the biggest issue
>is of course viewports, but since that's coming by DARPA, there's stuff
>like copy-on-capture and chdir to a non-executable file and syscalls.
>But nothing mission critical or half-broken. Right?
>  
>
We do the core traditional fs stuff, which is far less than being a 
general purpose namespace.

>  
>
>>Maybe filling out an app for grant would be good.  Is there an available
>>non-profit wrapper around Reiser4?  There must be some sort of shell for
>>the DARPA funding to have happened.
>>    
>>
>
>Non-profit wrapper? You mean some non-profit "welfare" organization?
>Why can't DARPA fund a company?
>  
>
They do.;-)

Unfortunately the state department won't let my team into the US, and 
this particular (SBIR) darpa grant requires the work be done in the US.

Also, it starts slowly requiring my personal time heavily, pays not a 
lot to me, and then, what I just found out, after 5 months of doing the 
exploratory work, you have a 2/3 chance of not getting funded to do the 
serious funding phase of the work.

I will never apply for an SBIR again I think.  I cannot afford to waste 
5 months of design work and then not get funded, these grants are for 
students and new professors.

Oh, and the EU doesn't care much to fund non-europeans.

It all involves so much more than just doing good work and getting 
paid.... whine whine.;-)

Re: a hosed reiserfs

[mjt]

On Wed, Jul 07, 2004 at 12:08:38AM -0700, Hans Reiser wrote:
>
>Unfortunately the state department won't let my team into the US, and 
>this particular (SBIR) darpa grant requires the work be done in the US.

What does SBIR stand for, btw?

That state department thing just plain sucks, damnit.

>Also, it starts slowly requiring my personal time heavily, pays not a 
>lot to me, and then, what I just found out, after 5 months of doing the 
>exploratory work, you have a 2/3 chance of not getting funded to do the 
>serious funding phase of the work.

The exploratory work does not include actual code yet?

So this means in essence that the viewport security is penalty-boxed?
(or what do you call it, delayed for the time being?)

>I will never apply for an SBIR again I think.  I cannot afford to waste 
>5 months of design work and then not get funded, these grants are for 
>students and new professors.

But at least the viewport security design exists and it's a damn
good idea, so I hope it will somehow get in at some point...

>Oh, and the EU doesn't care much to fund non-europeans.

Have you tried other nations around?
Although few have as much money as the United States.

>It all involves so much more than just doing good work and getting 
>paid.... whine whine.;-)

This whine is completely justified in my book ;)

-- 
mjt

Re: a hosed reiserfs

[Philippe Gramoullé]

Hello,

Feature wise, the lack of reiser4 quota support is definitely a showstopper for us, and maybe for a lot of people too.
So currently i'm focusing on performance of Reiser4 with databases, mainly MySQL.

Thanks

Philippe

--
Lycos Europe


On Wed, 7 Jul 2004 08:54:20 +0300
mjt@nysv.org (Markus Törnqvist) wrote:

  | Since feature-wise I don't see Reiser4 that lacking. Maybe the biggest issue
  | is of course viewports, but since that's coming by DARPA, there's stuff
  | like copy-on-capture and chdir to a non-executable file and syscalls.
  | But nothing mission critical or half-broken. Right?

Re: a hosed reiserfs

[Hans Reiser]

Philippe Gramoullé wrote:

>Hello,
>
>Feature wise, the lack of reiser4 quota support is definitely a showstopper for us, and maybe for a lot of people too.
>So currently i'm focusing on performance of Reiser4 with databases, mainly MySQL.
>
>Thanks
>
>Philippe
>
>--
>Lycos Europe
>
>
>On Wed, 7 Jul 2004 08:54:20 +0300
>mjt@nysv.org (Markus Törnqvist) wrote:
>
>  | Since feature-wise I don't see Reiser4 that lacking. Maybe the biggest issue
>  | is of course viewports, but since that's coming by DARPA, there's stuff
>  | like copy-on-capture and chdir to a non-executable file and syscalls.
>  | But nothing mission critical or half-broken. Right?
>
>
>  
>
Sigh.  Yes, if someone would sponsor quotas, it would be nice.  Thanks 
for pointing this out Philippe, it is a good candidate for what should 
be sponsored by someone looking to sponsor something.

Hans

Re: a hosed reiserfs

[mjt]

On Wed, Jul 07, 2004 at 12:18:05PM -0700, Hans Reiser wrote:

>Sigh.  Yes, if someone would sponsor quotas, it would be nice.  Thanks 
>for pointing this out Philippe, it is a good candidate for what should 
>be sponsored by someone looking to sponsor something.

Why not put up some estimates on cost for these features on the website?

Along with more WorldPay links, like "sponsor now" or "participate in
our hat-passing-along-for-quotas campaing" or something :)

I'm unfortunately in no position to sponsor really, but still some people
are and given some heads-up would probably not hurt...

Just a thought

Keep up the fight!

-- 
mjt

Re: a hosed reiserfs

[Hans Reiser]

Markus Törnqvist wrote:

>On Wed, Jul 07, 2004 at 12:08:38AM -0700, Hans Reiser wrote:
>  
>
>>Unfortunately the state department won't let my team into the US, and 
>>this particular (SBIR) darpa grant requires the work be done in the US.
>>    
>>
>
>What does SBIR stand for, btw?
>
>That state department thing just plain sucks, damnit.
>
>  
>
>>Also, it starts slowly requiring my personal time heavily, pays not a 
>>lot to me, and then, what I just found out, after 5 months of doing the 
>>exploratory work, you have a 2/3 chance of not getting funded to do the 
>>serious funding phase of the work.
>>    
>>
>
>The exploratory work does not include actual code yet
>  
>
Yes, it is supposed to show a "result", which seems bogus, so we are 
just going to pseudo-code all the algorithms, and then code a few 
performance critical ones and prove that they can be done, which is 
legitimate but..... They should just give us the full amount or not give 
it to us, nobody can do real work in 5 months.

>So this means in essence that the viewport security is penalty-boxed?
>(or what do you call it, delayed for the time being?)
>  
>
Well, it is waiting for the government to send us a contract to sign, 
etc., etc., and it will be as much of a burden as a blessing.  It will 
force me to hire some US folks.

>  
>
>>I will never apply for an SBIR again I think.  I cannot afford to waste 
>>5 months of design work and then not get funded, these grants are for 
>>students and new professors.
>>    
>>
>
>But at least the viewport security design exists and it's a damn
>good idea, so I hope it will somehow get in at some point...
>
>  
>
>>Oh, and the EU doesn't care much to fund non-europeans.
>>    
>>
>
>Have you tried other nations around?
>Although few have as much money as the United States.
>  
>
There is the EU and the US and nobody else sloshes much money around.  
Oh wait, maybe the Chinese.  I don't expect to get money from the 
Chinese as I suspect they are also nationalistic in giving it out.

>  
>
>>It all involves so much more than just doing good work and getting 
>>paid.... whine whine.;-)
>>    
>>
>
>This whine is completely justified in my book ;)
>
>  
>
thanks.

Re: a hosed reiserfs

[Philip Miller]

Hans Reiser wrote:
> Well, it is waiting for the government to send us a contract to sign, 
> etc., etc., and it will be as much of a burden as a blessing.  It will 
> force me to hire some US folks.

Is it just a matter of having US folks, or does a proportion of the money 
have to be paid to American employees?

Philip Miller

Re: a hosed reiserfs

[mjt]

On Thu, Jul 08, 2004 at 01:13:01PM -0400, Philip Miller wrote:
>
>Is it just a matter of having US folks, or does a proportion of the money 
>have to be paid to American employees?

Isn't it also that it's difficult for Russians to get employment visas
or such, permissions to leave the country to work elsewhere?

-- 
mjt

Re: a hosed reiserfs

[Hans Reiser]

Philip Miller wrote:

> Hans Reiser wrote:
>
>> Well, it is waiting for the government to send us a contract to sign, 
>> etc., etc., and it will be as much of a burden as a blessing.  It 
>> will force me to hire some US folks.
>
>
> Is it just a matter of having US folks, or does a proportion of the 
> money have to be paid to American employees?
>
> Philip Miller
>
>
work must be performed in the US.  Non-citizens can have to be approved 
but I imagine it won't be a problem.

Re: a hosed reiserfs

[Hans Reiser]

Markus Törnqvist wrote:

>On Thu, Jul 08, 2004 at 01:13:01PM -0400, Philip Miller wrote:
>  
>
>>Is it just a matter of having US folks, or does a proportion of the money 
>>have to be paid to American employees?
>>    
>>
>
>Isn't it also that it's difficult for Russians to get employment visas
>or such, permissions to leave the country to work elsewhere?
>
>  
>
that is exactly the problem.  These days work visas to the US for 
russians are very rare, not like a few years ago.

This causes me many problems.  I have other US customers besides darpa, 
and getting them comfortable with unseen workers is....
Hans