Large files under ubuntu do not appear to work

[don fisher]

Surfaz Gemon Meme wrote:
> 2009/3/22 don fisher <hdf3 at comcast.net <mailto:hdf3 at comcast.net>>
> 
>     Surfaz Gemon Meme wrote:
> 
>         2009/3/22 don fisher <hdf3 at comcast.net <mailto:hdf3 at comcast.net>
>         <mailto:hdf3 at comcast.net <mailto:hdf3 at comcast.net>>>
> 
> 
>            I am running ubuntu 8.10 on a Tyan workstation with 16 cores,
>         32GB of
>            ram and a 3Ware 8.9TB raid 5 array.
> 
> 
>         And your filesystem is...?
> 
>      >From /var/log/messages:
> 
>     sd 0:0:0:0: [sda] Very big device. Trying to use READ CAPACITY(16).
>     sd 0:0:0:0: [sda] Write Protect is off
>     sd 0:0:0:0: [sda] Write cache: enabled, read cache: disabled,
>     doesn't support DPO or FUA
>     sd 0:0:0:0: [sda] Very big device. Trying to use READ CAPACITY(16).
>     sd 0:0:0:0: [sda] 17577996288 512-byte hardware sectors (8999934 MB)
>     sd 0:0:0:0: [sda] Write Protect is off
>     sd 0:0:0:0: [sda] Write cache: enabled, read cache: disabled,
>     doesn't support DPO or FUA sda: sda1
>     sd 0:0:0:0: [sda] Attached SCSI disk
> 
>     3w-9xxx: scsi0: ERROR: (0x03:0x0101): Invalid command
>     opcode:opcode=0x4D.
>     3w-9xxx: scsi0: ERROR: (0x03:0x0101): Invalid command
>     opcode:opcode=0x4D.
> 
>     I will try do decode the error. Who is sending the command?
> 
>     I am using the driver that is part of the standard kernel. That is
>     also the one I used under Fedora.
> 
> 
> Please, send your e-mails to ubuntu-devel-discuss 
> <ubuntu-devel-discuss at lists.ubuntu.com 
> <mailto:ubuntu-devel-discuss at lists.ubuntu.com>>, not only to me.
> 

Sorry, I was not paying attention to my addresses. Today I started over, 
and was able to generate file systems up to 9TB. I have discovered that 
the damage is done during the reboot. I was successful rebooting with 
1TB and 2TB file systems. A 4TB file system failed the reboot. As I 
recall, it came up at 880MB.

I am reformatting the raid array, which takes about 24 hours. Any 
suggestions on what experiments I could perform would be appreciated. In 
my last position I had a 5.2TB array on a 4 core Opteron system using 
the same series of 3ware controllers, 9500S. So I do not think it is a 
controller problem.

don
-