#577608 IBM's Journaled File System (JFS) operation (mkdir) failed on huge logical volume

Package:
jfsutils
Source:
jfsutils
Description:
utilities for managing the JFS filesystem
Submitter:
"Feng Gao"
Date:
2010-04-13 03:15:04 UTC
Severity:
normal
#577608#5
Date:
2010-04-13 03:04:12 UTC
From:
To:
uname ¨Ca
Linux RANC01 2.6.26-2-amd64 #1 SMP Tue Jan 12 22:12:20 UTC 2010 x86_64
GNU/Linux

I successfully created the JFS file system on the huge logical volume
(consists of two equal-sized RAID 6 virtual disks, totally about 48T) and
mounted the volume. When I tried to make directory on it, it complained
about file io error. The following is excerpted from kern.log:

...
Mar 30 12:00:33 RANC01 kernel: [ 6758.339211] scsi 4:2:0:0: Direct-Access
DELL     PERC 6/E Adapter 1.22 PQ: 0 ANSI: 5
Mar 30 12:00:33 RANC01 kernel: [ 6758.373254] ses 4:0:17:0: Attached
Enclosure device
Mar 30 12:00:33 RANC01 kernel: [ 6758.373254] ses 4:0:17:0: Attached scsi
generic sg0 type 13
Mar 30 12:00:33 RANC01 kernel: [ 6758.373254] ses 4:0:34:0: Attached
Enclosure device
Mar 30 12:00:33 RANC01 kernel: [ 6758.373254] ses 4:0:34:0: Attached scsi
generic sg1 type 13
Mar 30 12:00:33 RANC01 kernel: [ 6758.373254] sd 4:2:0:0: [sda] Very big
device. Trying to use READ CAPACITY(16).
Mar 30 12:00:33 RANC01 kernel: [ 6758.373269] sd 4:2:0:0: [sda] 109366476800
512-byte hardware sectors (55995636 MB)
Mar 30 12:00:33 RANC01 kernel: [ 6758.373286] sd 4:2:0:0: [sda] Write
Protect is off
Mar 30 12:00:33 RANC01 kernel: [ 6758.373286] sd 4:2:0:0: [sda] Mode Sense:
1f 00 10 08
Mar 30 12:00:33 RANC01 kernel: [ 6758.373314] sd 4:2:0:0: [sda] Write cache:
disabled, read cache: enabled, supports DPO and FUA
Mar 30 12:00:33 RANC01 kernel: [ 6758.373338] sd 4:2:0:0: [sda] Very big
device. Trying to use READ CAPACITY(16).
Mar 30 12:00:33 RANC01 kernel: [ 6758.373352] sd 4:2:0:0: [sda] 109366476800
512-byte hardware sectors (55995636 MB)
Mar 30 12:00:33 RANC01 kernel: [ 6758.373365] sd 4:2:0:0: [sda] Write
Protect is off
Mar 30 12:00:33 RANC01 kernel: [ 6758.373365] sd 4:2:0:0: [sda] Mode Sense:
1f 00 10 08
Mar 30 12:00:33 RANC01 kernel: [ 6758.373393] sd 4:2:0:0: [sda] Write cache:
disabled, read cache: enabled, supports DPO and FUA
Mar 30 12:00:33 RANC01 kernel: [ 6758.373393]  sda:
Mar 30 12:00:33 RANC01 kernel: [ 6758.439120] sd 4:2:0:0: [sda] Attached
SCSI disk
Mar 30 12:00:33 RANC01 kernel: [ 6758.439120] sd 4:2:0:0: Attached scsi
generic sg2 type 0
Mar 30 12:02:39 RANC01 kernel: [ 6904.599049] ERROR: (device sda1):
dbAllocNext: Corrupt dmap page

The JFS file system was ok when only one RAID 6 (about 24T) was included in
logical volume.

Now, I use XFS file system, no problem at all.

regards,
Feng Gao