* Re: Desaster Reiser
2002-09-05 14:53 Desaster Reiser jose
@ 2002-09-05 10:09 ` Oleg Drokin
2002-09-05 11:04 ` Hans Reiser
[not found] ` <3D78CC6C.C09012AE@tridentemonigotes.com>
0 siblings, 2 replies; 6+ messages in thread
From: Oleg Drokin @ 2002-09-05 10:09 UTC (permalink / raw)
To: jose; +Cc: reiser
Hello!
On Thu, Sep 05, 2002 at 10:53:35AM -0400, jose@tridentemonigotes.com wrote:
> -I have a server 2.4.3 and something HD reiserfs. Yesterday one
> of this
> HD started to give me problems this is the message.
2.4.3 kernel is very old ad is filled with known bugs both in reiserfs and other
parts of code. You should consider upgrading.
> After this with reiserfs --check /dev/hdb2 this is the message that gave
> me
What reiserfsprogs version do you have?
If that's not the last one (3.6.3), you should download reiserfsprogs-3.6.3 from
our ftp site.
> I probed with badblocks and reiserfsck --rebuild -sd but was not
> anithing problem and my fear is I wrote that somebody have lose their HD
> with reiserfsck --rebuild -tree because after can't mount their units.I
That was possible because of reiserfsck crash for example, data are still not
lost though partition cannot be mounted.
> think that this is the last option if I have another
> Have another options before my HD is totally crash and all information
> is missing?
Well, if you want to be on safe side, you can create metadata snapshot
(debugreiserfs -p /dev/device | gzip -9c >metadata.gz) and make it available for
us to download. and we can verify that reiserfsprogs will work as expected.
Make sure you use reiserfsprogs 3.6.3 version for that operation.
Or you can download reiserfsprogs 3.6.3 and run reiserfsck --rebuild-tree
from there and only contact us if it will crash. (with metaadta, too).
Also after your problems are over you will need to upgrade the kernel to avoid
future corruptions.
Bye,
Oleg
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Desaster Reiser
2002-09-05 10:09 ` Oleg Drokin
@ 2002-09-05 11:04 ` Hans Reiser
2002-09-05 16:32 ` Rafael E. Herrera
[not found] ` <3D78CC6C.C09012AE@tridentemonigotes.com>
1 sibling, 1 reply; 6+ messages in thread
From: Hans Reiser @ 2002-09-05 11:04 UTC (permalink / raw)
To: Oleg Drokin; +Cc: jose, reiser
Oleg Drokin wrote:
>Hello!
>
>On Thu, Sep 05, 2002 at 10:53:35AM -0400, jose@tridentemonigotes.com wrote:
>
>
>> -I have a server 2.4.3 and something HD reiserfs. Yesterday one
>>of this
>>HD started to give me problems this is the message.
>>
>>
>
>2.4.3 kernel is very old ad is filled with known bugs both in reiserfs and other
>parts of code. You should consider upgrading.
>
>
>
>>
>>
Let's put this a bit more strongly. 2.4.3 should never have shipped to
users. It particularly has a ridiculously easy to hit bug in VFS that
affected reiserfs that would have been found if there was a policy of
test before shipping VFS changes. No one should be using it, and it is
definitely worth your time to upgrade.
We also have some not so easy but still possible to hit bugs in reiserfs
up until 2.4.18. I suggest 2.4.19/18 which is known to be rock stable
for us.
2.4.20-pre2 has some substantial performance improvements, but is more
risky as it has not been as well tested.
Hans
^ permalink raw reply [flat|nested] 6+ messages in thread
* Desaster Reiser
@ 2002-09-05 14:53 jose
2002-09-05 10:09 ` Oleg Drokin
0 siblings, 1 reply; 6+ messages in thread
From: jose @ 2002-09-05 14:53 UTC (permalink / raw)
To: reiser
Hello:
-I have a server 2.4.3 and something HD reiserfs. Yesterday one
of this
HD started to give me problems this is the message.
kernel: vs-4080:reiserfs-free-block:free-block (0342:137252)
(dev:blocknr); bit already cleared.
kernel: vs-13042: reiserfs_read_inode 2: (190310 190324 0x0 SD) not
found
kernel:vs-13048: reiserfs_iget: bad inode stat data of (190310 190324)
not found
After this with reiserfs --check /dev/hdb2 this is the message that gave
me
bad_leaf: block 8391 has invalid item 11: 2 42 0x1 DIR, len 452, entry
count 17, fsck need 0,
format old
block 198889 is not marked as used in the disk bitmap
block 442285 is not marked as used in the disk bitmap
block 442286 is not marked as used in the disk bitmap
block 442287 is not marked as used in the disk bitmap
block 442317 is not marked as used in the disk bitmap
block 442318 is not marked as used in the disk bitmap
block 442319 is not marked as used in the disk bitmap
grow_id_map: objectid map expanded: used 5120, 5 blocks
grow_id_map: objectid map expanded: used 5120, 5 blocks
grow_id_map: objectid map expanded: used 5120, 5 blocks
grow_id_map: objectid map expanded: used 5120, 5 blocks
grow_id_map: objectid map expanded: used 5120, 5 blocks
bad_leaf: block 394534 has invalid item 0: 197285 203828 0x73b8ad80 DIR,
len 112, entry count
4, fsck need 0, format old
free block count 413799 mismatches with a correct one 413826.
on-disk bitmap does not match to the correct one. 22 bytes differ
check_semantic_tree: hash mismatch detected (FactEmit.MYD)
check_semantic_tree: name "FactEmit.MYD" in directory 2 42 points to
nowhere
check_semantic_tree: hash mismatch detected (FactEmit.frm)
check_semantic_tree: name "FactEmit.frm" in directory 2 42 points to
nowhere
I intented debugreiserfs /dev/hdb2 and this is the out file
Super block of format 3.5 found on the 0x3 in block 16
Block count 4650656
Blocksize 4096
Free blocks 413799
Busy blocks (skipped 16, bitmaps - 142, journal blocks - 8193
1 super blocks, 4228505 data blocks
Root block 8403
Journal block (first) 18
Journal dev 0
Journal orig size 8192
Filesystem state ERROR
Tree height 4
Hash function used to sort names: "r5"
Objectid map size 250, max 1004
I probed with badblocks and reiserfsck --rebuild -sd but was not
anithing problem and my fear is I wrote that somebody have lose their HD
with reiserfsck --rebuild -tree because after can't mount their units.I
think that this is the last option if I have another
Have another options before my HD is totally crash and all information
is missing?
Thank Your for all and I don't hesitate.
Best Regards
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Desaster Reiser
2002-09-05 11:04 ` Hans Reiser
@ 2002-09-05 16:32 ` Rafael E. Herrera
2002-09-05 20:02 ` Hans Reiser
0 siblings, 1 reply; 6+ messages in thread
From: Rafael E. Herrera @ 2002-09-05 16:32 UTC (permalink / raw)
To: reiser
Hans Reiser wrote:
> Let's put this a bit more strongly. 2.4.3 should never have shipped to
> users. It particularly has a ridiculously easy to hit bug in VFS that
> affected reiserfs that would have been found if there was a policy of
> test before shipping VFS changes. No one should be using it, and it is
> definitely worth your time to upgrade.
>
> We also have some not so easy but still possible to hit bugs in reiserfs
> up until 2.4.18. I suggest 2.4.19/18 which is known to be rock stable
> for us.
> 2.4.20-pre2 has some substantial performance improvements, but is more
> risky as it has not been as well tested.
I would suggest you add a statement like the above in your web site.
--
Rafael
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Desaster Reiser
2002-09-05 16:32 ` Rafael E. Herrera
@ 2002-09-05 20:02 ` Hans Reiser
0 siblings, 0 replies; 6+ messages in thread
From: Hans Reiser @ 2002-09-05 20:02 UTC (permalink / raw)
To: Rafael E. Herrera; +Cc: reiser
Rafael E. Herrera wrote:
> Hans Reiser wrote:
>
>> Let's put this a bit more strongly. 2.4.3 should never have shipped
>> to users. It particularly has a ridiculously easy to hit bug in VFS
>> that affected reiserfs that would have been found if there was a
>> policy of test before shipping VFS changes. No one should be using
>> it, and it is definitely worth your time to upgrade.
>>
>> We also have some not so easy but still possible to hit bugs in
>> reiserfs up until 2.4.18. I suggest 2.4.19/18 which is known to be
>> rock stable for us. 2.4.20-pre2 has some substantial performance
>> improvements, but is more risky as it has not been as well tested.
>
>
> I would suggest you add a statement like the above in your web site.
We had one, but I decided that there were probably not a lot of remaing
2.4.3 users anymore after enough time went by.
Use 2.4.18 or later kernels. Marcelo is doing a great job and things
are stable by the time they stop having "pre" in their name.
^ permalink raw reply [flat|nested] 6+ messages in thread
* Re: Desaster Reiser
[not found] ` <3D78CC6C.C09012AE@tridentemonigotes.com>
@ 2002-09-06 16:06 ` jose
0 siblings, 0 replies; 6+ messages in thread
From: jose @ 2002-09-06 16:06 UTC (permalink / raw)
To: green; +Cc: reiser
jose@tridentemonigotes.com wrote:
>
> Oleg Drokin wrote:
> >
> > Hello!
> >
> > On Thu, Sep 05, 2002 at 10:53:35AM -0400, jose@tridentemonigotes.com wrote:
> > > -I have a server 2.4.3 and something HD reiserfs. Yesterday one
> > > of this
> > > HD started to give me problems this is the message.
> >
> > 2.4.3 kernel is very old ad is filled with known bugs both in reiserfs and other
> > parts of code. You should consider upgrading.
> >
> > > After this with reiserfs --check /dev/hdb2 this is the message that gave
> > > me
> >
> > What reiserfsprogs version do you have?
> > If that's not the last one (3.6.3), you should download reiserfsprogs-3.6.3 from
> > our ftp site.
> >
> > > I probed with badblocks and reiserfsck --rebuild -sd but was not
> > > anithing problem and my fear is I wrote that somebody have lose their HD
> > > with reiserfsck --rebuild -tree because after can't mount their units.I
> >
> > That was possible because of reiserfsck crash for example, data are still not
> > lost though partition cannot be mounted.
> >
> > > think that this is the last option if I have another
> > > Have another options before my HD is totally crash and all information
> > > is missing?
> >
> > Well, if you want to be on safe side, you can create metadata snapshot
> > (debugreiserfs -p /dev/device | gzip -9c >metadata.gz) and make it available for
> > us to download.
>
> I attach the file hdb2.gz after I create metadata snapshot. I hope that
> you can verify without any problem
>
> and we can verify that reiserfsprogs will work as expected.
> > Make sure you use reiserfsprogs 3.6.3 version for that operation.
>
> My old version of reiserfsprog was 3.6.2 but now is 3.6.3
>
> > Or you can download reiserfsprogs 3.6.3 and run reiserfsck --rebuild-tree
> > from there and only contact us if it will crash. (with metaadta, too).
>
> Respect this I would like another options before if that posible
> >
> > Also after your problems are over you will need to upgrade the kernel to avoid
> > future corruptions.
>
> You sure that after this problem I will upgrading and check the units
> every week
> >
> >Thank You
> By
>
> Jose
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2002-09-06 16:06 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2002-09-05 14:53 Desaster Reiser jose
2002-09-05 10:09 ` Oleg Drokin
2002-09-05 11:04 ` Hans Reiser
2002-09-05 16:32 ` Rafael E. Herrera
2002-09-05 20:02 ` Hans Reiser
[not found] ` <3D78CC6C.C09012AE@tridentemonigotes.com>
2002-09-06 16:06 ` jose
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.