Mike,
On Mon, 6 Aug 2007, Michael Hannon wrote:
> Greetings. We've got a new Dell PowerEdge (PE) 2950 server with a Dell
> MD 1000 external storage array attached. The PE 2950 is running
> Scientific Linux 5.0 with all patches installed:
>
>
> [root@ucdsys ~]# cat /etc/redhat-release
> Scientific Linux SL release 5.0 (Boron)
>
> [root@ucdsys ~]# cat /proc/version
> Linux version 2.6.18-8.1.8.el5 ([log in to unmask]) (gcc
> version 4.1.1 20070105 (Red Hat 4.1.1-52)) #1 SMP
> Thu Jul 12 15:15:31 EDT 2007
>
> The MD 1000 is attached via a PERC 5/E controller. The MD 1000 has 15
> 750GB SATA disk drives, arranged into two virtual disks, each containing
> 7 physical disks (with one global hot spare). The size of each virtual
> disk is 4.089TB. Please see the appended for more details.
>
> The two virtual drives appear in, say, fdisk as sdb and sdc, both of
> size 4497.6 GB. I'm a little confused by the fdisk output, which does
> show the 4GB size in the header but shows 2147480811 blocks as the size
> of the one and only partition. Again, see the appended for details.
>
> We have a problem in that we're unable to make the operating system see
> more than 2TB in these disks when we make ext3 file systems and mount
> the devices.
>
> The wikipedia article about ext3 claims that there are size limitations
> but that with a block size of 4096 (the default), the limits are 2TB per
> file and 16 TB per file system.
>
> Evidently I'm missing something basic here. Can anybody enlighten me?
You can't have partitions > 2 TB with the good old PC partition
tables fdisk handles. Instead, create a GPT label and your partition with
parted:
# parted /dev/sdb
(parted) mklabel gpt
(parted) mkpart primary 0 -0
(parted) q
You could probably use the device w/o partitions, but there may be a
reason why TUV doesn't support such setups.
- Stephan
> Thanks.
>
> - Mike
>
> Some details from the PERC 5/E set-up program
> ---------------------------------------------
>
> PERC 5/E
>
> Disk Group 0
>
> Virtual Disk 0
> RAID Level 5
> Size: 4.089TB
>
> Physical Disks
> 00 714880MB
> .
> .
> .
> 06
>
>
>
> Disk Group 1
>
> Virtual Disk 1
> RAID Level 5
> Size: 4.089TB
>
> Physical Disks
> 07 714880MB
> .
> .
> .
> 13
>
>
> ###### fdisk seems to see 4TB per device
>
> [root@ucdsys ~]# fdisk -l
> .
> .
> .
> Disk /dev/sdb: 4497.6 GB, 4497636065280 bytes
> 255 heads, 63 sectors/track, 546806 cylinders
> Units = cylinders of 16065 * 512 = 8225280 bytes
>
> Device Boot Start End Blocks Id System
> /dev/sdb1 1 267349 2147480811 83 Linux
>
> Disk /dev/sdc: 4497.6 GB, 4497636065280 bytes
> 255 heads, 63 sectors/track, 546806 cylinders
> Units = cylinders of 16065 * 512 = 8225280 bytes
>
> Device Boot Start End Blocks Id System
> /dev/sdc1 1 267349 2147480811 83 Linux
>
> ###### put a file system on the first device (4kb block size)
>
> [root@ucdsys ~]# mke2fs -j /dev/sdb1
> mke2fs 1.39 (29-May-2006)
> Filesystem label=
> OS type: Linux
> Block size=4096 (log=2)
> Fragment size=4096 (log=2)
> 268435456 inodes, 536870202 blocks
> 26843510 blocks (5.00%) reserved for the super user
> First data block=0
> Maximum filesystem blocks=4294967296
> 16384 block groups
> 32768 blocks per group, 32768 fragments per group
> 16384 inodes per group
> Superblock backups stored on blocks:
> 32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632,
> 2654208,
> 4096000, 7962624, 11239424, 20480000, 23887872, 71663616,
> 78675968,
> 102400000, 214990848, 512000000
>
> Writing inode tables: done
> Creating journal (32768 blocks): done
> Writing superblocks and filesystem accounting information: done
>
> This filesystem will be automatically checked every 24 mounts or
> 180 days, whichever comes first. Use tune2fs -c or -i to override.
>
> ###### Mount the new file system and examine its size
>
> [root@ucdsys ~]# mount /dev/sdb1 /backup1
>
> [root@ucdsys ~]# df -h /backup1
> Filesystem Size Used Avail Use% Mounted on
> /dev/sdb1 2.0T 199M 1.9T 1% /backup1
>
> ###### Only 2TB! But fdisk does report
>
> [root@ucdsys ~]# fdisk /dev/sdb
>
> The number of cylinders for this disk is set to 546806.
> There is nothing wrong with that, but this is larger than 1024,
> and could in certain setups cause problems with:
> 1) software that runs at boot time (e.g., old versions of LILO)
> 2) booting and partitioning software from other OSs
> (e.g., DOS FDISK, OS/2 FDISK)
>
> Command (m for help): p
>
> Disk /dev/sdb: 4497.6 GB, 4497636065280 bytes
> 255 heads, 63 sectors/track, 546806 cylinders
> Units = cylinders of 16065 * 512 = 8225280 bytes
>
> Device Boot Start End Blocks Id System
> /dev/sdb1 1 267349 2147480811 83 Linux
>
>
>
> Excerpt from wikipedia article about ext3:
> -----------------------------------------
> Size limits
>
> ext3 has a relatively small maximum size for both individual files and
> the
> entire filesystem. These limits are dependent on the block size of the
> filesystem; the following chart summarizes the limits[5]:
>
> Block size Max file size Max filesystem size
> 1KiB 16GiB 2TiB
> 2KiB 256GiB 8TiB
> 4KiB 2TiB 16TiB
> 8KiB 16TiB 32TiB
>
> The 8KiB block size is only available on architectures (such as alpha)
> which
> allow 8 KiB pages.
>
--
Stephan Wiesand
DESY - DV -
Platanenallee 6
15738 Zeuthen, Germany
|