Help with Raid5 mdadm Ubuntu 22.04

Operating Systems & Open Source Linux
,
1

Hi guys, i have been having some issues with create a raid5 using mdadm.

I got some used wd red drives from ebay. I plugged them via sata did mdadm. When i would do a reboot to make sure the raid has mounted properly, I would get a recovery mode cycle. When i did the windows checkdisk command it said it had some back sectors in them, I did a “fix” reformatted them into ext4, put them back into linux, created another array, it didnt stick recovering mode boot cycle happened again. Got tired and returned the disks.

Now, I bought some cheap Sandisk usbs to test to make sure im not going crazy and doing the install mdadm correctly (before install new out of the box drives) when i run the command " mdadm --detail /dev/md0 " the, everything seems correct except the state which says : “active,degraded,recovering” whichh does seem correct because these are brand spanking new usbs A 2.0. The state of the wd red drives said the same thing "active.degraded,recovering " does that imply its something else?

Also the USBs have been partitioned to use 60 out of 64GB. 3 drives in RAID 5.

Yes they are plugged directly on to the motherboard (in USB A 3.x), the motherboard does not have any usb A 2.0 on the rear io :frowning:

Guys what the heck im doing wrong, i am pretty close to using UNRAID but this should work, I have followed multiple tuts from jeffgeerling digital ocean etc

Also when i do the " sudo mdadm --detail–scan >> /etc/mdadm/mdadm.conf" it adds the UUID of the sdc, not md0 which i find odd, in yt tutorials they UUID is the md0 drive. I have done multiple clean installs so not sure whats happening there

computer specs:

i759-60x

970 gpu

ddr4 16gb 2133 crucial

nvme boot drive

cant find raid error on drives
cant find raid error on drives1920×1080 124 KB

Shell commands only cant find raid md0
Shell commands only cant find raid md01920×1080 208 KB

2

Welcome to Level1!

You do realize that a raid array first needs to sync, meaning read/write across the full drives. This process takes a couple of hours on hdds. I assume that you didn’t wait for the raid sync to finish.

As a result, the array will try to “recover” after every boot.

This process doesn’t change no matter if the drives are connected via internal SATA or USB.

You can check the progress of the array sync with

$ cat /proc/mdstat
3

That DEPEND says that your array wasn’t ready. Like @jode says, unless you overwrote the old disks with zeroes, the parity calculation will not match the existing data blocks on the disk. So the active/degraded/recovering can happen when an array is new and (in your RAID5 case) the parity stripe needs to be recalculated and rewritten, and will halt a boot because the all the mounts have to be ready as a stepping stone to a fully working installation. While it’s still recovering this, set your /etc/fstab to use option noauto for your array to get you into a non-recovery system.

K3n.