Skip to contentSkip to conversation

Can't mount after mkfs -b 8192

DISCUSSION David Greaves 3 participants 5 messages Jun 3, 2004 – Jun 4, 2004 MD RAID

lore
Share & tools
lore
#0 lore raw

Hi

First - thanks Guy - I've setup using 1 big partitions and I'll deal with monitoring as you suggest.

(BTW, are patches to mdadm recvd well? If so I'll take a look - especially at the state=spare message when rebuilding.)

Anyway, I now have an lvm2 on top and wanted a reiserfs (actually, since I'm ex-sgi I wanted xfs - but it doesn't shrink :( ) and I'm having odd behaviour. (Also, I posted to the reiser list but thinking about it, raid makes as much sense...)

So, environment:
Kernel 2.6.6
progsreiserfs 0.3.0.4-3

Summary:
I have a raid5 array in an LVM2 VG with a 600Gb LV defined. If I make a reiserfs with a block size of 8192 it won't mount.

The following commands were issued consecutively:

cu:~# mkfs -treiserfs /dev/video_vg/video_lv
All data on /dev/video_vg/video_lv will be lost. Do you really want to
create reiser filesystem (v3.6) (y/n) y
Creating reiser filesystem (v3.6) with standard journal on
/dev/video_vg/video_lv
initializing skiped area: done
initializing journal: done
syncing...done
cu:~# mount /dev/video_vg/video_lv /huge
cu:~# df -k
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/hda2             19283776    697988  18585788   4% /
/dev/hda1                97826     13003     79604  15% /boot
/dev/mapper/video_vg-video_lv
                    629126396     32840 629093556   1% /huge
cu:~# umount /huge
cu:~# mkfs -treiserfs -b 8192 /dev/video_vg/video_lv
All data on /dev/video_vg/video_lv will be lost. Do you really want to
create reiser filesystem (v3.6) (y/n) y
Creating reiser filesystem (v3.6) with standard journal on
/dev/video_vg/video_lv
initializing skiped area: done
initializing journal: done
syncing...done
cu:~# mount /dev/video_vg/video_lv /huge
mount: wrong fs type, bad option, bad superblock on /dev/video_vg/video_lv,
      or too many mounted file systems

dmesg now shows:
sh-2011: read_super_block: can't find a reiserfs filesystem on (dev dm-0, block 16, size 4096)
sh-2021: reiserfs_fill_super: can not find reiserfs on dm-0

any suggestions.

Nb I'll be storing 2-5Gb video files - hence the large blocksize.

David

#1 lore raw
Quoted 7 lines · On Thursday June 3, [email protected] wrote:

Hi

First - thanks Guy - I've setup using 1 big partitions and I'll deal with monitoring as you suggest.

(BTW, are patches to mdadm recvd well? If so I'll take a look - especially at the state=spare message when rebuilding.)

I like to think so.....

NeilBrown

#2 lore raw
In reply to David Greaves

I am sure a patch would be considered! But what do I know?
I do agree that the state should be something other than "clean, no-errors"! Maybe:
"clean, no-errors, degraded"
Or
"clean, no-errors, degraded, recovering"

Not even sure it's clean if a disk is missing. The parity can't be up to date!
Also seems like a missing disk would be an error!

Maybe need a list of possible states and the meanings to understand.

Don't know about your mount issue, but I did notice 4096 in you dmesg, should it be 8192? I did not see any mount options for block size.

sh-2011: read_super_block: can't find a reiserfs filesystem on (dev dm-0, block 16, size 4096)

Guy

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of David Greaves
Sent: Thursday, June 03, 2004 6:32 PM
To: [email protected]
Subject: Can't mount after mkfs -b 8192

Hi

First - thanks Guy - I've setup using 1 big partitions and I'll deal with monitoring as you suggest.

(BTW, are patches to mdadm recvd well? If so I'll take a look - especially at the state=spare message when rebuilding.)

Anyway, I now have an lvm2 on top and wanted a reiserfs (actually, since I'm ex-sgi I wanted xfs - but it doesn't shrink :( ) and I'm having odd behaviour. (Also, I posted to the reiser list but thinking about it, raid makes as much sense...)

So, environment:
Kernel 2.6.6
progsreiserfs 0.3.0.4-3

Summary:
I have a raid5 array in an LVM2 VG with a 600Gb LV defined. If I make a reiserfs with a block size of 8192 it won't mount.

The following commands were issued consecutively:

cu:~# mkfs -treiserfs /dev/video_vg/video_lv
All data on /dev/video_vg/video_lv will be lost. Do you really want to
create reiser filesystem (v3.6) (y/n) y
Creating reiser filesystem (v3.6) with standard journal on
/dev/video_vg/video_lv
initializing skiped area: done
initializing journal: done
syncing...done
cu:~# mount /dev/video_vg/video_lv /huge
cu:~# df -k
Filesystem           1K-blocks      Used Available Use% Mounted on
/dev/hda2             19283776    697988  18585788   4% /
/dev/hda1                97826     13003     79604  15% /boot
/dev/mapper/video_vg-video_lv
                     629126396     32840 629093556   1% /huge
cu:~# umount /huge
cu:~# mkfs -treiserfs -b 8192 /dev/video_vg/video_lv
All data on /dev/video_vg/video_lv will be lost. Do you really want to
create reiser filesystem (v3.6) (y/n) y
Creating reiser filesystem (v3.6) with standard journal on
/dev/video_vg/video_lv
initializing skiped area: done
initializing journal: done
syncing...done
cu:~# mount /dev/video_vg/video_lv /huge
mount: wrong fs type, bad option, bad superblock on /dev/video_vg/video_lv,
       or too many mounted file systems

dmesg now shows:
sh-2011: read_super_block: can't find a reiserfs filesystem on (dev dm-0, block 16, size 4096)
sh-2021: reiserfs_fill_super: can not find reiserfs on dm-0

any suggestions.

Nb I'll be storing 2-5Gb video files - hence the large blocksize.

David

-
To unsubscribe from this list: send the line "unsubscribe linux-raid" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#3 lore raw
Quoted 6 lines · On Thursday June 3, [email protected] wrote:

I am sure a patch would be considered! But what do I know?
I do agree that the state should be something other than "clean, no-errors"! Maybe:
"clean, no-errors, degraded"
Or
"clean, no-errors, degraded, recovering"

Good suggestion. It will be in the next release of mdadm (due out shortly).
The "no-errors" will be gone though. It doesn't mean anything.

Quoted 3 lines

Not even sure it's clean if a disk is missing. The parity can't be up to date!

Well.... the parity had better be up-to-date, or you have data corruption.
When an array is running, the kernel will maintain sufficient state that it always knows what should be where, but the data actually on disk may be "clean" or not.
If you crash with a non-clean array with a failed disk, then you loose data (and need --assemble --force to restart the array).

Quoted 1 line

Also seems like a missing disk would be an error!

No, a degradation. 2 missing disks in a raid5 would be a failure. I don't report that yet.

NeilBrown

#4 lore raw
In reply to Guy
Quoted 7 lines · Guy

Don't know about your mount issue, but I did notice 4096 in you dmesg, should it be 8192? I did not see any mount options for block size.

sh-2011: read_super_block: can't find a reiserfs filesystem on (dev dm-0, block 16, size 4096)

Yep
I got a response on the reiser list that prompted me to try XFS.

It rather nicely said:
XFS: Attempted to mount file system with blocksize 8192 bytes
XFS: Only page-sized (4096) or less blocksizes currently work.
XFS: SB validate failed

David
PS http://www.namesys.com/mkreiserfs.html says :
-b | --block-size N
N is block size in bytes. It may only be set to a power of 2 within
the 512-8192 interval.

So that might help googlers in the future ;)