#27 Beyond end of device accesses occurring

open
nobody
None
5
2005-08-26
2005-08-25
Daniel D Parry
No

Hello, I wonder if anyone can shed some light on a
problem I have encountered? I am using
linux-iscsi-5.0.0.0.3rc6-363 and a 2.6.11.1 kernel. I
find I can attach and mount the device /dev/sdb1
successfully:

Log messages:
iscsid: version 0.3 variant (22-Apr-2005)
Vendor: IFT Model: A12E-G2121-2 Rev: 342D
Type: Direct-Access ANSI SCSI
revision: 03
SCSI device sdb: 2048000000 512-byte hdwr sectors
(1048576 MB)
SCSI device sdb: drive cache: write through
SCSI device sdb: 2048000000 512-byte hdwr sectors
(1048576 MB)
SCSI device sdb: drive cache: write through
sdb: sdb1
Attached scsi disk sdb at scsi0, channel 0, id 0, lun 1

df output:
Filesystem Size Used Avail Use% Mounted on
/dev/sdb1 962G 9.3G 904G 2% /mnt/sanstorage

fdisk -l output:
Disk /dev/sdb: 1048.5 GB, 1048576000000 bytes
255 heads, 63 sectors/track, 127482 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id
System
/dev/sdb1 1 127482 1023999133+ 83
Linux

I am using it a storage point for backuppc. So when I
start backuppc, data is effectively being almost
continuously written to the device when the backups are
being run. However, after a short while, when say
around 5GB of data has been transferred, I encounter
the following kind of errors:

attempt to access beyond end of device
sdb1: rw=0, want=27426812160, limit=2047998267

Shortly after this, I get:

EXT3-fs error (device sdb1): ext3_free_blocks: Freeing
blocks not in datazone - block = 3428351519, count = 1
Aborting journal on device sdb1
ext3_abort called.
EXT3-fs error (device sdb1): ext3_journal_start_sb:
Detected aborted journal
Remounting filesystem read-only
EXT3-fs error (device sdb1) in
ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device sdb1) in ext3_truncate: Journal
has aborted
EXT3-fs error (device sdb1) in
ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device sdb1) in ext3_orphan_del: Journal
has aborted
EXT3-fs error (device sdb1) in
ext3_reserve_inode_write: Journal has aborted
EXT3-fs error (device sdb1) in ext3_delete_inode:
Journal has aborted
EXT3-fs error (device sdb1) in start_transaction:
Journal has aborted
EXT3-fs error (device sdb1) in start_transaction:
Journal has aborted
__journal_remove_journal_head: freeing b_committed_data
EXT3-fs error (device sdb1) in start_transaction:
Journal has aborted

So I'm wondering why the beyond end of device accesses
are occurring in the first instance? If anyone has any
suggestions, they would be most appreciated!

Discussion

  • Daniel D Parry
    Daniel D Parry
    2005-08-25

    • status: open --> closed
     
  • Daniel D Parry
    Daniel D Parry
    2005-08-25

    Logged In: YES
    user_id=1120997

    Aha. I think this may have been caused by file system
    errors, possibly resulting from when I accidentally had two
    machines with identical initiator names accessing the SAN.
    I've run fsck on the 1TB device (^_^), and repaired, and
    things seem ok, thus far, after about 9GB transferred...

    Will close this support request. Apologies, for wasting any
    time, though may reopen if the problem returns... hopefully
    it won't!

     
  • Logged In: YES
    user_id=1221775

    Dan, I'm curious. How did you get two servers with the same
    iqn's? I had entcountered this mysterious issue once in our
    QA group and chalked it up to an operator error because the
    engineer was never able to reproduce it.
    Regards,
    Wayne.

     
  • Daniel D Parry
    Daniel D Parry
    2005-08-25

    Logged In: YES
    user_id=1120997

    Just for the record, shortly after my last post...

    ...it fell over (>_<) My plan now is to fdisk the device
    again, and rebuild the filesystem from scratch...

    Oh, and Wayne I just had two linux boxes with identical
    /etc/initiatorname.iscsi files, both running iscsid and
    attached to the SAN. This caused problems... (^_^)

     
  • Daniel D Parry
    Daniel D Parry
    2005-08-26

    Logged In: YES
    user_id=1120997

    Spoke much too soon... am still encountering this problem,
    even after a resh fdiisk and mkfs. Any ideas...?

     
  • Daniel D Parry
    Daniel D Parry
    2005-08-26

    • status: closed --> open