[SOLVED] Home server with raid0 and raid1 issue after reboot

karcio · 09-17-2020, 03:36 PM

Hi Guys,

Trying to set home server on centos8. I've installed os on raid0 with 2x120GB ssd. Also I did setup raid1 2x500GB disks.
So after that all was correct

Code:

[root@localhost karcio]# df -h
Filesystem      Size  Used Avail Use% Mounted on
devtmpfs        7.3G     0  7.3G   0% /dev
tmpfs           7.3G     0  7.3G   0% /dev/shm
tmpfs           7.3G  8.8M  7.3G   1% /run
tmpfs           7.3G     0  7.3G   0% /sys/fs/cgroup
/dev/md126       49G  1.8G   45G   4% /
/dev/sda2       976M  141M  769M  16% /boot
/dev/md125      151G   61M  143G   1% /home
/dev/sda1       599M  6.9M  592M   2% /boot/efi
tmpfs           1.5G     0  1.5G   0% /run/user/1000
/dev/md0        458G   73M  434G   1% /mnt/raid1

my fstab looks like

Code:

UUID=59c2c137-52eb-4842-9c63-8d788e33484b /                       ext4    defaults        1 1
UUID=9e6f2deb-411d-4277-bdd1-08dbd46fa338 /boot                   ext4    defaults        1 2
UUID=BED2-6553          /boot/efi               vfat    umask=0077,shortname=winnt 0 2
UUID=54a02e7b-f6a8-4a75-bd17-2de277b1ba15 /home                   ext4    defaults        1 2
UUID=34b70f95-9b7b-471b-be5f-92381f5b59cc swap                    swap    defaults        0 0
UUID=d92f879d-d0c0-44c8-bd7a-1a7a534609c4    /mnt/raid1   ext4   defaults   0   0

but after reboot I can see that:
- mount /mnt/raid1 is failing
- dependency failed for /home
- dependency failed for Local File system

and system does not start. I guess I did somewhere mistake, not sure where

Please have a look in screenshot

https://i.postimg.cc/MT9NZH9T/photo-2020-09-17-21-21-35.jpg

lvm_ · 09-18-2020, 03:01 AM

What is in your /etc/mdadm/mdadm.conf file? What is the contents of /proc/mdstat? What happens when you start the array manually? (also you have three, not two arrays) What happens when you run 'mdadm -s -A'?

karcio · 09-18-2020, 10:03 AM

Hi please have a look on results

Quote:

What happens when you start the array manually?

-

https://i.postimg.cc/gkL9sqZd/photo5...8708264979.jpg

https://i.postimg.cc/fWmX41Xr/photo5...8708264977.jpg

https://i.postimg.cc/28JVVw4Z/photo5...8708264976.jpg

berndbausch · 09-18-2020, 03:32 PM

I'd also like to see the result of lsblk -f and/or blkid if these commands are available.

lvm_ · 09-19-2020, 04:47 AM

Your mdadm.conf is a mess: multiple lines for same arrays, missing swap. I recommend re-creating it from scratch, could be the reason why the system is not booting. Start with basic

Code:

DEVICE partitions
HOMEHOST <system>
MAILADDR root

run 'mdadm --detail --scan >> /etc/mdadm/mdadm.conf' to add array descriptions and edit array names if needed. Ok, backup the original file first.

Also, I made a mistake: 'mdadm -A -s', not 'mdadm -s -A' - sorry. If it start all arrays it means that arrays are fine but configuration files are not.

syg00 · 09-19-2020, 05:07 AM

Root on RAID0 - seriously ?. See my sigline.

karcio · 09-19-2020, 05:31 AM

Hi please have a look in commands output

https://i.postimg.cc/W4pN4DtK/photo5...8521950172.jpg

https://i.postimg.cc/c1Pf8WRT/photo5...8521950171.jpg

Looks like I have duplicated MD device names in config file, for create file I was using follow command

Quote:

mdadm --detail --scan --verbose | sudo tee -a /etc/mdadm.conf

please notice that config file has different location in centos8

@syg00 what do you mean about root on raid0, is it wrong practice? Please explain, this is my first attempt to raid, so if I doing something wrong please explain what, or share a link I can read,

Thank you all for your help

syg00 · 09-19-2020, 05:50 AM

Read the first couple of sentences of this - you cannot afford the risk; the root doesn't need any purported speed increase RAID0 supposedly offers.

lvm_ · 09-19-2020, 06:07 AM

Every practice has its place. RAID0 is fast and has no overhead but also has no redundancy and will fail if any of the disk fail, it is quite all right to use it if performance is essential while reliability is not and you don't mind restoring non-booting system from a backup now and then. Not something I would do though.

karcio · 09-21-2020, 04:06 PM

Hi guys, thank you for all your suggestion.

recreate mdadm config file help me and machine could boot

Code:

mdadm --detail --scan >> /etc/mdadm.conf

https://i.postimg.cc/pTHGSGCN/photo-...1-22-01-31.jpg

but after I've read https://en.wikipedia.org/wiki/Standa..._levels#RAID_0

Decided not to do raid0 on my 2x120GB disks. Instead this I did LVM.

Now it is looks like:

Code:

[root@localhost karcio]# df -h
Filesystem           Size  Used Avail Use% Mounted on
devtmpfs             7.3G     0  7.3G   0% /dev
tmpfs                7.3G     0  7.3G   0% /dev/shm
tmpfs                7.3G  8.8M  7.3G   1% /run
tmpfs                7.3G     0  7.3G   0% /sys/fs/cgroup
/dev/mapper/cl-root   49G  1.8G   45G   4% /
/dev/sda2            976M  143M  766M  16% /boot
/dev/mapper/cl-home  153G   61M  145G   1% /home
/dev/sda1            599M  6.9M  592M   2% /boot/efi
tmpfs                1.5G     0  1.5G   0% /run/user/1000
/dev/md0             458G   73M  434G   1% /mnt/raid1

Regards
karcio