[CentOS] 2 TB limit on USB drive

Mon Nov 16 20:12:56 UTC 2009
Gareth Tupper <gareth at phonepower.com>

Hallo
 
I submitted this as a bug several weeks ago, but I wanted to ask around
& see if anyone else has come across this....
 
	I have a USB Buffalo Drivestation Quattro, with 4 1TB disks
configured in raid5 as one 2.8TB (or so) disk, attached to a Cent 5.4 64
bit server (completely yum'd up to date)

The disk is labeled as GPT, and formatted as a 2.8 TB ext3 partition
(this issue also happens with xfs).  I used a gparted boot disk to
create the partition.

When I attach the drive I see this in messages:

Nov 2 14:26:55 kernel: usb 1-5.2: new high speed USB device using
ehci_hcd and address 7
Nov 2 14:26:56 kernel: usb 1-5.2: configuration 0000001
<http://bugs.centos.org/view.php?id=1>  chosen from 1 choice
Nov 2 14:26:56 kernel: scsi5 : SCSI emulation for USB Mass Storage
devices
Nov 2 14:26:56 kernel: usb-storage: device found at 7
Nov 2 14:26:56 kernel: usb-storage: waiting for device to settle before
scanning
Nov 2 14:27:01 kernel: Vendor: BUFFALO Model: HD-QSSU2/R5 1 Rev: 2.02
Nov 2 14:27:01 kernel: Type: Direct-Access ANSI SCSI revision: 05
Nov 2 14:27:01 kernel: sdc : very big device. try to use READ
CAPACITY(16).
Nov 2 14:27:01 kernel: sdc : READ CAPACITY(16) failed.
Nov 2 14:27:01 kernel: sdc : status=0, message=00, host=5, driver=00
Nov 2 14:27:01 kernel: sdc : use 0xffffffff as device size
Nov 2 14:27:01 kernel: SCSI device sdc: 4294967296 512-byte hdwr sectors
(2199023 MB)
Nov 2 14:27:01 kernel: sdc: Write Protect is off

After this failure, the disk is either a) inaccessible, or b) reports
only a 2 TB partition.

The latest Ubuntu can read the disk, presenting the full 2.8 TB just
peachy.

This server is up to date:
uname -a: Linux myserver.mydomain.com 2.6.18-164.6.1.el5 #1 SMP Tue Nov
3 16:18:27 EST 2009 i686 i686 i386 GNU/Linux
cat /etc/redhat-release: CentOS release 5.4 (Final)

[root at myserver ~]# cat /proc/partitions
major minor #blocks name
   ...
   8 32 2147483648 sdc << the disk showing incorrectly with only 2TB of
storage
 
This bug seems very similar to a previous bug:
https://bugzilla.redhat.com/show_bug.cgi?id=502944 which was reported
fixed in 5.4 
 
Anyone seen this before, or have any ideas how I can get CentOS to see
the disk?
 
 
 
Cheers,
Gareth
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.centos.org/pipermail/centos/attachments/20091116/29396cde/attachment-0004.html>