Hello everyone. I’m trying to install CentOS 8 with root and swap partitions on software raid. The plan is:
– create md0 raid level 1 with 2 hard drives: /dev/sda and /dev/sdb, using Linux Rscue CD,
– install CentOS 8 with Virtual Box on my laptop,
– rsync CentOS 8 root partition on /dev/md0p1,
– chroot in CentOS 8 root partition,
– configure /etc/mdadm.conf, grub.cfg, initramfs, install bootloader on both sda and sdb drives.

I think I can do first four of the above, but my CentOS installation acts strange after rebooting the server. It recognizes the raid, but boots randomly with root on /dev/sda1 (and recognizes raid with /dev/sdb disk), or with root on /dev/sdb1 (and recognizes raid with /dev/sda disk). When booting from Linux Rescue CD, the raid with two disk is recognized.

Creating md0 wth Linux Rescue CD.

# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 819.5M 1 loop /run/archiso/sfs/airootfs sda 8:0 0 232.9G 0 disk sdb 8:16 0 232.9G 0 disk sr0 11:0 1 871M 0 rom

# mdadm –create /dev/md0 –level=1 –raid-devices=2
–metadata=0.9 /dev/sda /dev/sdb mdadm: array /dev/md0 started.

# mdadm –detail –scan –verbose ARRAY /dev/md0 level=raid1 num-devices=2 metadata=0.90 UUID=fef2f68b:
1a462c80:c44c77eb:7ee19756 devices=/dev/sda,/dev/sdb

# cat /proc/mdstat Personalities : [raid1]
md0 : active raid1 sdb[1] sda[0]
244198464 blocks [2/2] [UU]
[>………………..] resync = 1.9% (4682944/244198464)
finish=52.2min speed=76356K/sec bitmap: 2/2 pages [8KB], 65536KB chunk

unused devices:

(1 hour later)

# cat /proc/mdstat Personalities : [raid1]
md0 : active raid1 sdb[1] sda[0]
244198464 blocks [2/2] [UU]
bitmap: 0/2 pages [0KB], 65536KB chunk

unused devices:

# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 819.5M 1 loop /run/archiso/sfs/airootfs sda 8:0 0 232.9G 0 disk
└─md0 9:0 0 232.9G 0 raid1
sdb 8:16 0 232.9G 0 disk
└─md0 9:0 0 232.9G 0 raid1
sr0 11:0 1 871M 0 rom

———————————

So far so good. Now partitions:

# fdisk /dev/md0

Welcome to fdisk (util-linux 2.33.1). Changes will remain in memory only, until you decide to write them. Be careful before using the write command.

Device does not contain a recognized partition table. Created a new DOS disklabel with disk identifier 0x64998720.

Command (m for help): o Created a new DOS disklabel with disk identifier 0xfe784f57.

Command (m for help): n Partition type
p primary (0 primary, 0 extended, 4 free)
e extended (container for logical partitions)
Select (default p): p Partition number (1-4, default 1):
First sector (2048-488396927, default 2048):
Last sector, +/-sectors or +/-size{K,M,G,T,P} (2048-488396927, default
488396927): +220G

Created a new partition 1 of type ‘Linux’ and of size 220 GiB.

Command (m for help): n Partition type
p primary (1 primary, 0 extended, 3 free)
e extended (container for logical partitions)
Select (default p): p Partition number (2-4, default 2):
First sector (461375488-488396927, default 461375488):
Last sector, +/-sectors or +/-size{K,M,G,T,P} (461375488-488396927, default 488396927): +4G

Created a new partition 2 of type ‘Linux’ and of size 4 GiB.

Command (m for help): w The partition table has been altered. Calling ioctl() to re-read partition table. Syncing disks.

# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
loop0 7:0 0 819.5M 1 loop /run/archiso/sfs/airootfs sda 8:0 0 232.9G 0 disk
└─md0 9:0 0 232.9G 0 raid1
├─md0p1 259:0 0 220G 0 part
└─md0p2 259:1 0 4G 0 part
sdb 8:16 0 232.9G 0 disk
└─md0 9:0 0 232.9G 0 raid1
├─md0p1 259:0 0 220G 0 part
└─md0p2 259:1 0 4G 0 part
sr0 11:0 1 871M 0 rom

# mkfs -t ext4 /dev/md0p1
mke2fs 1.44.5 (15-Dec-2018)
Creating filesystem with 57671680 4k blocks and 14417920 inodes Filesystem UUID: 511d0d07-4d55-4b6c-939e-d23ae2da1ce0
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632,
2654208, 4096000, 7962624, 11239424, 20480000, 23887872

Allocating group tables: done
Writing inode tables: done
Creating journal (262144 blocks): done Writing superblocks and filesystem accounting information: done

# mkswap /dev/md0p2
Setting up swapspace version 1, size = 4 GiB (4294963200 bytes)
no label, UUID=5667d007-0d3e-4b54-8a72-6ade52609adc

————————————–

Mount would be root partition:

# mkdir /mnt/CentOS8
# mount -t ext4 /dev/md0p1 /mnt/CentOS8/

————————————–

OK, I have what I need. Now log into virtualised CentOS 8 in Virtual Box.

# rpm -q rsync rsync-3.1.3-7.el8.x86_64
# rpm -q mdadm mdadm-4.1-13.el8.x86_64

# rsync -vaxAXHSz –delete / 192.168.111.167:/mnt/CentOS8/
The authenticity of host ‘192.168.111.167 (192.168.111.167)’ can’t be established. ECDSA key fingerprint is SHA256:HukswAQlarthYqbpVHIHm80JaWHKX12AxgOn8UuKQxk. Are you sure you want to continue connecting (yes/no/[fingerprint])? yes Warning:
Permanently added ‘192.168.111.167’ (ECDSA) to the list of known hosts. root@192.168.111.167’s password: sending incremental file list
./
bin -> usr/bin lib -> usr/lib lib64 -> usr/lib64
sbin -> usr/sbin
[lots and lots of files…]
var/yp/

sent 694,980,592 bytes received 745,802 bytes 1,795,422.95 bytes/sec total size is 1,558,934,091 speedup is 2.24

————————————–

Chrooting time:

# mount –types proc /proc /mnt/CentOS8/proc
# mount –rbind /sys /mnt/CentOS8/sys
# mount –make-rslave /mnt/CentOS8/sys
# mount –rbind /dev /mnt/CentOS8/dev
# mount –make-rslave /mnt/CentOS8/dev
# chroot /mnt/CentOS8 /usr/bin/bash
# source /etc/profile

Configure raid:

# mdadm –detail –scan –verbose ARRAY /dev/md0 level=raid1 num-devices=2 metadata=0.90 UUID=fef2f68b:
1a462c80:c44c77eb:7ee19756 devices=/dev/sda,/dev/sdb
# mdadm –detail –scan –verbose > /etc/mdadm.conf

Edit /etc/fstab:

# cat /etc/fstab

#
# /etc/fstab
# Created by anaconda on Sun Nov 8 10:05:31 2020
#
# Accessible filesystems, by reference, are maintained under
# ‘/dev/disk/’. See man pages fstab(5), findfs(8), mount(8) and/or
# blkid(8) for more info.
#
# After editing this file, run ‘systemctl daemon-reload’ to update
# systemd units generated from this file.
#
#UUID=36c5a887-79c3-4b6a-b433-0e4d7574b1e4 /
#ext4 defaults 1 1
/dev/md0p1 / ext4 defaults 1 1

Configure and install bootloader:

# cat /etc/default/grub GRUB_TIMEOUT=8
GRUB_DISTRIBUTOR=”$(sed ‘s, release .*$,,g’ /etc/system-release)”
GRUB_DEFAULT=saved GRUB_DISABLE_SUBMENU=true GRUB_TERMINAL_OUTPUT=”console”
GRUB_CMDLINE_LINUX=”rhgb quiet”
GRUB_DISABLE_RECOVERY=”true”
GRUB_ENABLE_BLSCFG=true

# grub2-mkconfig -o /boot/grub2/grub.cfg Generating grub configuration file … done

# grub2-install /dev/sda Installing for i386-pc platform. Installation finished. No error reported.
# grub2-install /dev/sdb Installing for i386-pc platform. Installation finished. No error reported.

Configure initramfs:

# dracut -f /boot/initramfs-4.18.0-193.28.1.el8_2.x86_64.img
# 4.18.0-193.28.1.el8_2.x86_64

——————————–

Reboot and login to my new CentOS 8 installation:

# lsblk NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINT
sda 8:0 0 232.9G 0 disk sdb 8:16 0 232.9G 0 disk
├─sdb1 8:17 0 220G 0 part /
└─sdb2 8:18 0 4G 0 part sr0 11:0 1 871M 0 rom

# mdadm –detail –scan –verbose INACTIVE-ARRAY /dev/md127 num-devices=1 metadata=0.90 UUID=fef2f68b:
1a462c80:c44c77eb:7ee19756 devices=/dev/sda

# cat /proc/mdstat Personalities :
md127 : inactive sda[0](S)
244198464 blocks

unused devices:

I can live with md127, because it happens constantly on multiple distributions, but why root is on /dev/sdb1 (and sometimes it’s /dev/sda1)? I’m running out of spells now. :) Do anyone can suggest what else I forgot to do?

thanks

—
Łukasz Posadowski