I had a md array check trigerred on my nas yesterday. It completed without errors.
I restared the system after apt upgrade.

Today I wanted to reboot again as I installed a new kernel. I systemd told
me something was prebenting to reboot. It was again due to an md array check.
cat /proc/mdstat
Personalities : [raid10] [linear] [raid0] [raid1] [raid6] [raid5] [raid4]
md0 : active raid10 sda1[6] sdd1[5] sdc1[0] sdf1[4] sde1[7](S)
      5860268032 blocks super 1.2 512K chunks 2 near-copies [4/4] [UUUU]
      [=====>...............]  check = 26.4% (1550913536/5860268032) finish=331.1min speed=216906K/sec

unused devices: <none>

I then tried to find why, and found that it now use systemd via timer instead of crontab.

/usr/lib/systemd/system/mdcheck_start.timer is the culprit: I checked the systax for once
a moth as before and found that the actual setting was not that:


OnCalendar=Sun *-*-1..7 1:00:00

should be changed to

OnCalendar=Sun *-*-* 1:00:00

See also

https://build.opensuse.org/projects/openSUSE:Step:15-SP2/packages/mdadm.15570/files/1002-OnCalendar-format-fix-of-mdcheck_start-timer.patch?expand=0

I checked the old syntax from source below and apparently it is correct.

See https://wiki.archlinux.org/title/Systemd/Timers.

Maybe it is due to systemd 258~rc3-1 update.