ok I think this is the right place for this as it is related to install process not data storage.
I just want to help any new user who want to run OMV on a raid-1 system disk.
I have found several very good help blogs describing the process of converting system to raided setup
blog #1 by Josip Lazić http://lazic.info/josip/post/i…ediavault-on-raid-device/
blog #2 by François Marier https://feeding.cloud.geek.nz/…ting-up-raid-on-existing/
I have found however that some steps do not work all the time or not as I was expecting them to, so with some trial and error here are exact steps that did worked perfectly for me.
!A WORD OF CAUTION !
please check each and every command for proper disk references.
I had to reinstall twice simply because I used wrong disk as target.
Assumptions taken
#0 we are using OMV version 2.+ (my actual version was 2.2.4)
#1 the initial setup is done as normal on /dev/sda
if you are installing a-new plugin only 2 drives you want to use as system drive when all is done.
in my case I used 2 120GB SSD
an Intel SSD 535 Series and a SanDisk SDSSDA12 series
I run install on /dev/sda (intel)
and used SanDisk to create mdadm raid for conversion.
SO
Step 1. install both disks in the machine. make sure they are the only drives installed and run OMV setup.
a. if you can use CD for installation. I had to scrap several setups do to "can not install grab" error when using USB stick.
yes even though I was installing on SSD it still tried to install grab on usb stick I was installing from and destroyed the stick.
b. choose to install to /dev/sda when warned about multiple drives found on the system. for some reason OMV prefers /sda over anything else so just use the KISS
c. boot the machine and configure with webui. change admin password what ever.
make sure everything is running as expected, configure network etc.
d. start ssh server.
Step 2. once all is working open terminal and ssh into machine
we can skip the installation of the tools as instructed in the blogs above since OMV already have all of that installed
A. Partition the empty drive (/dev/sdb). by default OMV install creates 3 partitions 1. boot(MBR) 2. / and 3 . swap
we will replicate the same.
from blog 1 : and I am using numbers for my config to set partition size
parted -a optimal /dev/sdb mklabel gpt
parted -a optimal /dev/sdb mkpart grub ext2 2048s 12M = 10M !! this is our boot partition !!
parted -a optimal /dev/sdb mkpart root ext4 12M 111GB = 103GB !! will be used for md0 "/" !!
parted -a optimal /dev/sdb mkpart data ext4 111GB 100% = 8.4GB!! md1 "swap" !!
parted -a optimal /dev/sdb set 1 bios_grub on
parted -a optimal /dev/sdb set 2 boot on
this is the only thing I used from blog 1
the rest is from blog 2
B. Create new degraded mdadm array
NOTE: we need to change the lines below to point to /dev/sdb2 and /dev/sdb3 not 1 and 2 as /dev/sdb1 is our boot partition
mdadm --create /dev/md0 --level=1 --raid-devices=2 missing /dev/sdb2
mdadm --create /dev/md1 --level=1 --raid-devices=2 missing /dev/sdb3
C. format new raid devices
adjust /etc/mdadm/mdadm.conf by adding "DEVICE /dev/sda* /dev/sdb*
run GRUB config and choose both drives a and b to install grub to. Make sure you choose the /dev/sda and /dev/sdb nothing else
D. mount the new raid volume and copy existing drive to it
mkdir /tmp/mntroot
mount /dev/md0 /tmp/mntroot
rsync -auHxv --exclude=/proc/* --exclude=/sys/* --exclude=/tmp/* /* /tmp/mntroot/
E. run a blkid and get the uuid for md0 and md1
edit fstab in "/tmp/mntroot/etc/fstab" and change uuid for "/" and "swap" to the respected uuid for md0 and md1
Then reboot and from within the GRUB menu, hit "e" to enter edit mode and make sure that you will be booting off of the new disk:
set root='(md/0)'
linux /boot/vmlinuz-2.6.32-5-amd64 root=/dev/md0 ro quiet
and yes you can actually use the values seen here. the omv grab use hd0,mbr notation for root in set root, and uuid for root in second line just use /dev/md0
Once the system is up, you can check that the root partition is indeed using the RAID array by running mount and looking for something like:
"/dev/md0 on / type ext4 (rw,noatime,errors=remount-ro)"
now if all looks good use the same code we use on top to partition the sda drive, this will essentially wipe-out the drive and create exactly the same partiton table we hav on sdb drive
parted -a optimal /dev/sda mklabel gpt
parted -a optimal /dev/sda mkpart grub ext2 2048s 12M = 10M !! this is our boot partition !!
parted -a optimal /dev/sda mkpart root ext4 12M 111GB = 103GB !! will be used for md0 "/" !!
parted -a optimal /dev/sda mkpart data ext4 111GB 100% = 8.4GB!! md1 "swap" !!
parted -a optimal /dev/sda set 1 bios_grub on
parted -a optimal /dev/sda set 2 boot on
add new partiton to raid array and wait until it finish synchronizing
mdadm /dev/md0 -a /dev/sda2
mdadm /dev/md1 -a /dev/sda3
perform the test described by François Marier in the second blog to ensure all is as expected.
and we are done. we can not continue setting up the system as we like
good luck