I've done a little research, and can't really find out why it happens, but on one server I have built, I get the old fashion /dev/sda1, /dev/sda2 type partitions and on another server I get the /dev/mapper style partitions. Both machines are hardware raided (before the install) and both machines are Centos 5.5 64 bit, using the same set of CDs to install from.
Can someone please explain why this happens? I'm not sure it matters, so this is just a curious bother. The machines are different brands with different raid controllers. Both were partitoned using the "custom layout" when installed. I'm thinking I must have overlooked a checkbox or something during the install.
This is not a big deal, not a problem, it's just an itch in my brain.
Thanks for any explanation.
steve campbell
Am 24.06.2011 14:24, schrieb Steve Campbell:
I've done a little research, and can't really find out why it happens, but on one server I have built, I get the old fashion /dev/sda1, /dev/sda2 type partitions and on another server I get the /dev/mapper style partitions. Both machines are hardware raided (before the install) and both machines are Centos 5.5 64 bit, using the same set of CDs to install from.
Can someone please explain why this happens? I'm not sure it matters, so this is just a curious bother. The machines are different brands with different raid controllers. Both were partitoned using the "custom layout" when installed. I'm thinking I must have overlooked a checkbox or something during the install.
This is not a big deal, not a problem, it's just an itch in my brain.
Thanks for any explanation.
steve campbell
The machine with the /dev/sda has a true hardware RAID controller. The machine with a /dev/mapper device has just a fake RAID controller, thus a dm-raid device.
Alexander
Alexander Dalloz wrote:
Am 24.06.2011 14:24, schrieb Steve Campbell:
I've done a little research, and can't really find out why it happens, but on one server I have built, I get the old fashion /dev/sda1, /dev/sda2 type partitions and on another server I get the /dev/mapper style partitions. Both machines are hardware raided (before the install) and both machines are Centos 5.5 64 bit, using the same set of CDs to install from.
Can someone please explain why this happens? I'm not sure it matters, so this is just a curious bother. The machines are different brands with different raid controllers. Both were partitoned using the "custom layout" when installed. I'm thinking I must have overlooked a checkbox or something during the install.
This is not a big deal, not a problem, it's just an itch in my brain.
Thanks for any explanation.
steve campbell
The machine with the /dev/sda has a true hardware RAID controller. The machine with a /dev/mapper device has just a fake RAID controller, thus a dm-raid device.
Alexander
Instead of fake-raid controller, I suggest software RAID via mdadm (Linux md).
Ljubomir
At Fri, 24 Jun 2011 08:24:57 -0400 CentOS mailing list centos@centos.org wrote:
I've done a little research, and can't really find out why it happens, but on one server I have built, I get the old fashion /dev/sda1, /dev/sda2 type partitions and on another server I get the /dev/mapper style partitions. Both machines are hardware raided (before the install) and both machines are Centos 5.5 64 bit, using the same set of CDs to install from.
Same or different RAID controllers? Different RAID controller drivers present the logical disks to the O/S in different ways. *Most* present the logical disks as SCSI disks (with contrived vendor and model values) and others do something else (eg with their own block devices the old DAC960 Mylex driver is an example of the latter). Are both machines using LVM? Did you previously install some other operating system before you install CentOS on either machine? It could be that the machine with /dev/sda1, /dev/sda2 type partitions may have been previously partitioned or something. The 'default' installer setup on a completely bare system is for the installer to create two partitions, /dev/sda1 for /boot and /dev/sda2 for a LVM volume group, which is then broken in to a swap volume and a root volume.
Can someone please explain why this happens? I'm not sure it matters, so this is just a curious bother. The machines are different brands with different raid controllers. Both were partitoned using the "custom layout" when installed. I'm thinking I must have overlooked a checkbox or something during the install.
This is not a big deal, not a problem, it's just an itch in my brain.
Thanks for any explanation.
steve campbell
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
-
Alexander Dalloz -
Ljubomir Ljubojevic -
Robert Heller -
Steve Campbell