Hey there guys,
I’ve successfully setup a raid 0 with 2 of my NVMEs. The raid is working well (in terms of being able to be used).
However, I’ve observed a few things that made me reconsider if it indeed is working well.
I’ve setup the raid following this Arch Linux guide .
When I tried to do my first scrubbing via
echo check > /sys/block/md127/md/sync_action
it gave me a permission error:
LC_ALL=C echo check > /sys/block/md127/md/sync_action
zsh: keine Berechtigung: /sys/block/md127/md/sync_action
(I don’t know why LC_ALL=C isn’t working here, but nevermind, just telling me I don’t have permission).
3 things I’ve wondered now:
Why is it giving me permission errors?
Why does my raid have this arbitrary number 127 in the first place? Did I set up something wrong?
Why are the UUIDs mismatching from /etc/mdadm.conf and from blkid /dev/md127?
In the conf it says 63b5e92f:a8276105:cb8562bd:b69717d6 and blkid prints c8acd199-af63-4583-99a3-c9896b346668
Here’s an excerpt from my /etc/mdadm.conf:
ARRAY /dev/md/DownloadCache metadata=1.2 UUID=63b5e92f:a8276105:cb8562bd:b69717d6
My questions would now be: Is this expected or did I mess up at some point? If I did, how can I fix it? And also, why is the scrubbing command not working for me/printing no permissions?
Any help would be greatly appreciated
I also tried starting from scratch, which didn’t work as expected. I just can’t seem to entirely delete all traces of that raid.
I must’ve turned wrong direction at some point I can’t see right now
I remember setting up raid using md/raid - I made some notes - given the fact that the server is no longer in service - I don’t have more than those to offer.
1 Like
Thank you Aarhus once again for chiming in.
I’m onto something and will report back with my findings - and a possible solution/error analysis
Sadly neither my research nor your notes have helped me resolving the issue, sigh.
Still getting no permission error when running
sudo echo check > /sys/block/mdX/md/sync_action
xabbu
27 July 2024 06:13
6
You need root permissions to write to this sys file. Your command has not the appropriate permissions to write. A simple sudo echo cannot work, since the redirect is run with your user permissions.
For example check the file owner for the newly created file
sudo echo TEST > /tmp/testfile
ls -l /tmp/testfile
You should switch to your root account
su -l
and then try
echo check > /sys/block/mdX/md/sync_action
An alternativ is to use tee.
tee works fine, one of the simplest programs in your /usr/bin.
LC_ALL=C echo check | sudo tee /sys/block/md127/md/sync_action
2 Likes
Thank you for chiming in. Sadly, both commands are still giving me permission errors. I don’t know how I can get past this
LC_ALL=C echo check | sudo tee /sys/block/md0/md/sync_action
tee: /sys/block/md0/md/sync_action: Keine Berechtigung
check
[PowerTower ~]# echo check > /sys/block/md0/md/sync_action
-bash: /sys/block/md0/md/sync_action: Keine Berechtigung
Note that the raid device has changed from md127 to md0. I can’t seem to be able to start this task…
have you checked groups ?
Switch to root context before executing the command
su -l root
Then - you should make sure the block device is correct - simple copy/paste may not work
echo check > /sys/block/md127/md/sync_action
Is the md127 correct?
In all of the Arch wikis, the example commands are prepended with a prompt.
the # stands for the root accountsu to become root)
Just prepending sudo does not always work.
As you found out,sudo echo check > ...echo check > ... when issued from the root account
The intention ofLC_ALL=CLANG=C)
Some commands are chained together - the output of one is the input to the next
su -l root
Passwort:
[PowerTower ~]# echo check > /sys/block/md0/md/sync_action
-bash: /sys/block/md0/md/sync_action: Keine Berechtigung
[PowerTower ~]#
That’s what I did, but sadly it still doesn’t like it.
I did not. And I’m not sure how I would go about inspecting the groups. Here’s the id of my own user, is there anything peculiar?
id donatus
uid=1000(donatus) gid=1001(donatus) Gruppen=1001(donatus),998(wheel),991(lp),3(sys),90(network),98(power),1000(autologin),959(docker)
Check the raid build process
cat /proc/mdstat
You need the process to be finished before you can do anything with it.
It is finished I guess and it’s showing the raid as active:
cat /proc/mdstat
Personalities : [raid0]
md0 : active raid0 nvme1n1p1[1] nvme2n1p1[0]
1953257472 blocks super 1.2 512k chunks
unused devices: <none>
if you list your block devices you should get something like
[server1 ~]# lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sda 8:0 0 223,6G 0 disk
├─sda1 8:1 0 300M 0 part /boot/efi
├─sda2 8:2 0 214,5G 0 part /srv/samba/public
│ /srv/samba/data
│ /
└─sda3 8:3 0 8,8G 0 part [SWAP]
sdb 8:16 0 447,1G 0 disk
sdc 8:32 0 447,1G 0 disk
sdd 8:48 0 447,1G 0 disk
└─sdd1 8:49 0 400G 0 part
└─md127 9:127 0 399,9G 0 raid10
sde 8:64 0 447,1G 0 disk
└─sde1 8:65 0 400G 0 part
└─md127 9:127 0 399,9G 0 raid10
sdf 8:80 1 0B 0 disk
sdg 8:96 1 0B 0 disk
sdh 8:112 1 0B 0 disk
sdi 8:128 1 0B 0 disk
sdj 8:144 1 0B 0 disk
I do sir!
lsblk
NAME MAJ:MIN RM SIZE RO TYPE MOUNTPOINTS
sda 8:0 0 16,4T 0 disk
└─sda2 8:2 0 16,4T 0 part /mnt/disk1
sdb 8:16 0 16,4T 0 disk
└─sdb2 8:18 0 16,4T 0 part /mnt/disk2
sdc 8:32 0 20T 0 disk
├─sdc1 8:33 0 16M 0 part
└─sdc2 8:34 0 20T 0 part /mnt/disk3
sdd 8:48 1 0B 0 disk
sde 8:64 1 0B 0 disk
nvme1n1 259:0 0 931,5G 0 disk
└─nvme1n1p1 259:1 0 931,5G 0 part
└─md0 9:0 0 1,8T 0 raid0 /mnt/DownloadCache
nvme0n1 259:2 0 3,6T 0 disk
├─nvme0n1p1 259:3 0 300M 0 part /boot/efi
└─nvme0n1p2 259:4 0 3,6T 0 part /
nvme2n1 259:5 0 931,5G 0 disk
└─nvme2n1p1 259:6 0 931,5G 0 part
└─md0 9:0 0 1,8T 0 raid0 /mnt/DownloadCache
nvme3n1 259:7 0 3,6T 0 disk
└─nvme3n1p1 259:8 0 3,6T 0 part /mnt/EmbyData
Then if you want to scrub them you target the md0 device
As root
echo check > /sys/block/md0/md/sync_action
But…that’s what I’m doing already for the past hours, which sadly doesn’t work as I said earlier:
sudo -i
[sudo] Passwort für donatus:
[PowerTower ~]# echo check > /sys/block/md0/md/sync_action
-bash: /sys/block/md0/md/sync_action: Keine Berechtigung
That’s exactly where it always says “No permission”
Hmm - I missed that - sorry.
What happens if you use
sudo mdadm --action=check /dev/md0
But since the scrubbing is to locate disk errors and repair data before failure - raid0 is just speed - no redundancy - is it necessary in such case?
The raid I used was for redundancy and speed - raid10far2 - and check is necessary to preemptive locate data errors before they become critical.
No problem at all sir!
That’s exactly what I’ve tried earlier, too! And I was wondering why this doesn’t work either!
sudo mdadm --action=check /dev/md0
LC_ALL=C sudo mdadm --action=check /dev/md0echo
mdadm: Couldn't open /dev/md0echo: No such file or directory
Which is weird, as lsblk reports it as “/dev/md0”.
That’s an entirely reasonable question. It’s not that I think it is urgent or necessary, it’s just that I tried it earlier, but wasn’t able to ever execute it, so I began asking myself if I’ve even set up the entire raid correctly
