Hello Artur, Your answer worked! I just wanted to follow up, so anyone else who has a similar issue can arrive at the solution. Also, I am reporting what I did since it is not exactly as you suggested. I tried to follow your extra step, but it returned as an invalid number. I ran with sudo since my user is not root in this case: printf "%llu\n" -1 > sudo /sys/block/md126/md/resync_start -bash: printf: /sys/block/md126/md/resync_start: invalid number So, we assumed that you simply wanted us to edit the resync_start file value to the number 18446744073709551615. I did it by hand using a text editor. After doing so, the value of the file changed to none. sudo cat /sys/block/md126/md/resync_start none After that, I proceeded to reconstruct the array. But I changed the order of the commands. Not sure if that mattered. sudo mdadm -S /dev/md125 mdadm: stopped /dev/md125 sudo mdadm -a /dev/md127 /dev/sdb mdadm: added /dev/sdb sudo mdadm -R --force /dev/md126 mdadm: Started /dev/md/Data with 3 devices (0 new) Even though it only reported 3 devices (0 new) during the last command's output, it successfully added the new /dev/sdb drive as a spare, started the array resync, and is recovering now as reported by cat /proc/mdstat. Thank you so much for the assistance! Devon Beets From: Artur Paszkiewicz <artur.paszkiewicz@xxxxxxxxx> Sent: Wednesday, April 7, 2021 4:25 AM To: Devon Beets <devon@xxxxxxxxxxxxxxxx>; Tkaczyk, Mariusz <mariusz.tkaczyk@xxxxxxxxxxxxxxx>; linux-raid@xxxxxxxxxxxxxxx <linux-raid@xxxxxxxxxxxxxxx> Cc: Glenn Wikle <gwikle@xxxxxxxxxxxxxxxx> Subject: Re: Cannot add replacement hard drive to mdadm RAID5 array On 06.04.2021 22:05, Devon Beets wrote: > Output of the recommended commands for adding the new disk to the RAID5 array: > > sudo mdadm -R --force /dev/md126 > mdadm: array /dev/md/Data now has 3 devices (0 new) > > sudo mdadm -S /dev/md125 > mdadm: stopped /dev/md125 > > sudo mdadm -a /dev/md127 /dev/sdb > mdadm: added /dev/sdb > > Output of mdstat after running the commands. Shows that both md126 and md127 are inactive, and there is no RAID resync happening. > > cat /proc/mdstat > Personalities : [raid6] [raid5] [raid4] [linear] [multipath] [raid0] [raid1] [raid10] > md126 : inactive sda[2] sdc[1] sdd[0] > 5567507712 blocks super external:/md127/0 > > md127 : inactive sdb[3](S) sdc[2](S) sdd[1](S) sda[0](S) > 10564 blocks super external:imsm > > unused devices: <none> It looks like mdadm still does not handle this case correctly. Please do this before the "mdadm -R --force /dev/md126": printf "%llu\n" -1 > /sys/block/md126/md/resync_start Regards, Artur
