I'm playing with software RAID and LVM in some virtual machines and I've run into an issue that I can't find a good answer to in the docs.
I have the following RAID setup:
md0: sda1 and sdb1, RAID 1. This is /boot
md1: sda2 and sdb2, RAID 1. This is a PV for LVM.
VolGroup00, this is the volume group and md1 is the only PV in it.
LogVol00 is swap LogVol01 is / LogVol02 is /home
So, I tested to see what happens if I disable sdb in virtualbox. Machine booted find and I was able to see that part of the raid array was gone.
I reattached the disk and rebuilt the array mdam --add /dev/md0 /dev/sdb1 mdam --add /dev/md1 /dev/sdb2
The array rebuilt without issue. But now, if I remove sda, the machine doesn't boot and grub complains. The question I have is how does one reinstall the boot sector at this point? I'm a little confused as far as to what device to install it on and what to specify as the root.
This is grub.conf if it helps:
default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz title CentOS (2.6.18-194.e15) root (hd0,0) kernel /vmlinuz-2.6.18-194.e15 ro root=/dev/VolGroup00/LogVol01 initrd /initrd-2.6.18-194.e15.img
Grub has always been a little confusing for me, so I guess I don't understand. What do the two root entries specify? And again, my main question is how do I reinstall grub to the repaired disk? Sorry for the lengthy post, I'm trying to provide as much information as possible.
On 09/18/2010 05:13 PM, Matthew Topper wrote:
I'm playing with software RAID and LVM in some virtual machines and I've run into an issue that I can't find a good answer to in the docs.
I have the following RAID setup:
md0: sda1 and sdb1, RAID 1. This is /boot
md1: sda2 and sdb2, RAID 1. This is a PV for LVM.
VolGroup00, this is the volume group and md1 is the only PV in it.
LogVol00 is swap LogVol01 is / LogVol02 is /home
So, I tested to see what happens if I disable sdb in virtualbox. Machine booted find and I was able to see that part of the raid array was gone.
I reattached the disk and rebuilt the array mdam --add /dev/md0 /dev/sdb1 mdam --add /dev/md1 /dev/sdb2
The array rebuilt without issue. But now, if I remove sda, the machine doesn't boot and grub complains. The question I have is how does one reinstall the boot sector at this point? I'm a little confused as far as to what device to install it on and what to specify as the root.
This is grub.conf if it helps:
default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz title CentOS (2.6.18-194.e15) root (hd0,0) kernel /vmlinuz-2.6.18-194.e15 ro root=/dev/VolGroup00/LogVol01 initrd /initrd-2.6.18-194.e15.img
Grub has always been a little confusing for me, so I guess I don't understand. What do the two root entries specify? And again, my main question is how do I reinstall grub to the repaired disk? Sorry for the lengthy post, I'm trying to provide as much information as possible.
http://wiki.centos.org/HowTos/CentOS5ConvertToRAID#head-fa2b73a28acdf965daa1...
HTH,
Timo :)
Entering those commands exactly worked, but I'm not sure why it worked. So, it's progress.
I need to find some good reference on GRUB, which seems difficult because most things I can find are about GRUB2, and CentOS has 0.97
On Sat, 18 Sep 2010 17:42:59 +0200 Timo Schoeler timo.schoeler@riscworks.net wrote:
On 09/18/2010 05:13 PM, Matthew Topper wrote:
I'm playing with software RAID and LVM in some virtual machines and I've run into an issue that I can't find a good answer to in the docs.
I have the following RAID setup:
md0: sda1 and sdb1, RAID 1. This is /boot
md1: sda2 and sdb2, RAID 1. This is a PV for LVM.
VolGroup00, this is the volume group and md1 is the only PV in it.
LogVol00 is swap LogVol01 is / LogVol02 is /home
So, I tested to see what happens if I disable sdb in virtualbox. Machine booted find and I was able to see that part of the raid array was gone.
I reattached the disk and rebuilt the array mdam --add /dev/md0 /dev/sdb1 mdam --add /dev/md1 /dev/sdb2
The array rebuilt without issue. But now, if I remove sda, the machine doesn't boot and grub complains. The question I have is how does one reinstall the boot sector at this point? I'm a little confused as far as to what device to install it on and what to specify as the root.
This is grub.conf if it helps:
default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz title CentOS (2.6.18-194.e15) root (hd0,0) kernel /vmlinuz-2.6.18-194.e15 ro root=/dev/VolGroup00/LogVol01 initrd /initrd-2.6.18-194.e15.img
Grub has always been a little confusing for me, so I guess I don't understand. What do the two root entries specify? And again, my main question is how do I reinstall grub to the repaired disk? Sorry for the lengthy post, I'm trying to provide as much information as possible.
http://wiki.centos.org/HowTos/CentOS5ConvertToRAID#head-fa2b73a28acdf965daa1...
HTH,
Timo :) _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Sat, Sep 18, 2010 at 2:41 PM, Matthew Topper topperm9@gmail.com wrote:
On Sat, 18 Sep 2010 17:42:59 +0200 Timo Schoeler timo.schoeler@riscworks.net wrote:
On 09/18/2010 05:13 PM, Matthew Topper wrote:
I'm playing with software RAID and LVM in some virtual machines and I've run into an issue that I can't find a good answer to in the docs.
I have the following RAID setup:
md0: sda1 and sdb1, RAID 1. This is /boot
md1: sda2 and sdb2, RAID 1. This is a PV for LVM.
VolGroup00, this is the volume group and md1 is the only PV in it.
LogVol00 is swap LogVol01 is / LogVol02 is /home
So, I tested to see what happens if I disable sdb in virtualbox. Machine booted find and I was able to see that part of the raid array was gone.
I reattached the disk and rebuilt the array mdam --add /dev/md0 /dev/sdb1 mdam --add /dev/md1 /dev/sdb2
The array rebuilt without issue. But now, if I remove sda, the machine doesn't boot and grub complains. The question I have is how does one reinstall the boot sector at this point? I'm a little confused as far as to what device to install it on and what to specify as the root.
This is grub.conf if it helps:
default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz title CentOS (2.6.18-194.e15) root (hd0,0) kernel /vmlinuz-2.6.18-194.e15 ro root=/dev/VolGroup00/LogVol01 initrd /initrd-2.6.18-194.e15.img
Grub has always been a little confusing for me, so I guess I don't understand. What do the two root entries specify? And again, my main question is how do I reinstall grub to the repaired disk?
http://wiki.centos.org/HowTos/CentOS5ConvertToRAID#head-fa2b73a28acdf965daa1...
Entering those commands exactly worked, but I'm not sure why it worked. So, it's progress.
I need to find some good reference on GRUB, which seems difficult because most things I can find are about GRUB2, and CentOS has 0.97
The commands installed grub into the mbr of sdb that way you can boot from either sda or sdb if the other is missing.
You should always do that when booting from raid1 http://wiki.centos.org/HowTos/SoftwareRAIDonCentOS5
GRUB1 manual http://www.gnu.org/software/grub/manual/legacy/grub.html
Good link (AFAIR) http://members.iinet.net.au/~herman546/p15.html
Thanks. That third link especially helped.
On Sat, Sep 18, 2010 at 5:47 PM, Tom H tomh0665@gmail.com wrote:
On Sat, Sep 18, 2010 at 2:41 PM, Matthew Topper topperm9@gmail.com wrote:
On Sat, 18 Sep 2010 17:42:59 +0200 Timo Schoeler timo.schoeler@riscworks.net wrote:
On 09/18/2010 05:13 PM, Matthew Topper wrote:
I'm playing with software RAID and LVM in some virtual machines and I've run into an issue that I can't find a good answer to in the docs.
I have the following RAID setup:
md0: sda1 and sdb1, RAID 1. This is /boot
md1: sda2 and sdb2, RAID 1. This is a PV for LVM.
VolGroup00, this is the volume group and md1 is the only PV in it.
LogVol00 is swap LogVol01 is / LogVol02 is /home
So, I tested to see what happens if I disable sdb in virtualbox. Machine booted find and I was able to see that part of the raid array was gone.
I reattached the disk and rebuilt the array mdam --add /dev/md0 /dev/sdb1 mdam --add /dev/md1 /dev/sdb2
The array rebuilt without issue. But now, if I remove sda, the machine doesn't boot and grub complains. The question I have is how does one reinstall the boot sector at this point? I'm a little confused as far as to what device to install it on and what to specify as the root.
This is grub.conf if it helps:
default=0 timeout=5 splashimage=(hd0,0)/grub/splash.xpm.gz title CentOS (2.6.18-194.e15) root (hd0,0) kernel /vmlinuz-2.6.18-194.e15 ro root=/dev/VolGroup00/LogVol01 initrd /initrd-2.6.18-194.e15.img
Grub has always been a little confusing for me, so I guess I don't understand. What do the two root entries specify? And again, my main question is how do I reinstall grub to the repaired disk?
http://wiki.centos.org/HowTos/CentOS5ConvertToRAID#head-fa2b73a28acdf965daa1...
Entering those commands exactly worked, but I'm not sure why it worked. So, it's progress.
I need to find some good reference on GRUB, which seems difficult because most things I can find are about GRUB2, and CentOS has 0.97
The commands installed grub into the mbr of sdb that way you can boot from either sda or sdb if the other is missing.
You should always do that when booting from raid1 http://wiki.centos.org/HowTos/SoftwareRAIDonCentOS5
GRUB1 manual http://www.gnu.org/software/grub/manual/legacy/grub.html
Good link (AFAIR) http://members.iinet.net.au/~herman546/p15.html _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Mon, Sep 20, 2010 at 8:51 AM, Matthew Topper topperm9@gmail.com wrote:
On Sat, Sep 18, 2010 at 5:47 PM, Tom H tomh0665@gmail.com wrote:
I need to find some good reference on GRUB, which seems difficult because most things I can find are about GRUB2, and CentOS has 0.97
The commands installed grub into the mbr of sdb that way you can boot from either sda or sdb if the other is missing.
You should always do that when booting from raid1 http://wiki.centos.org/HowTos/SoftwareRAIDonCentOS5
GRUB1 manual http://www.gnu.org/software/grub/manual/legacy/grub.html
Good link (AFAIR) http://members.iinet.net.au/~herman546/p15.html
Thanks. That third link especially helped.
You're welcome.
-
Matthew Topper -
Timo Schoeler -
Tom H