On Sun, Mar 30, 2025 at 11:56 PM pgnd <pgnd@xxxxxxxxxxxx> wrote: > > on > > distro > Name: Fedora Linux 41 (Forty One) > Version: 41 > Codename: > > mdadm -V > mdadm - v4.3 - 2024-02-15 > > rpm -qa | grep mdadm > mdadm-4.3-4.fc41.x86_64 > > i have a relatively-new (~1 month) 4x4TB RAID10 array. > > after a reboot, one of the drives got kicked > > dmesg > ... > [ 15.513443] sd 15:0:7:0: [sdn] Attached SCSI disk > [ 15.784537] md: kicking non-fresh sdn1 from array! > ... > > cat proc mdstat > md124 : active raid10 sdm1[1] sdl1[0] sdk1[4] > 7813770240 blocks super 1.2 512K chunks 2 near-copies [4/3] [UU_U] > bitmap: 1/59 pages [4KB], 65536KB chunk > > smartctl shows no issues; can't yet find a reason for the kick. > > re-adding the drive, rebuild starts. > > it's progressing with recovery; after ~ 30mins, I see > > md124 : active raid10 sdm1[1] sdn1[2] sdl1[0] sdk1[4] > 7813770240 blocks super 1.2 512K chunks 2 near-copies [4/3] [UU_U] > [=========>...........] recovery = 49.2% (1924016576/3906885120) finish=3918230862.4min speed=0K/sec > bitmap: 1/59 pages [4KB], 65536KB chunk > > the values of > > finish=3918230862.4min speed=0K/sec > > appear nonsensical. > > is this a bug in my mdadm config, progress reporting & or an actual problem with _function_? > > Hi Are there some D state progress? And how about `ps auxf | grep md`? Is there a filesystem on it? If so, can you still read/write data from it? Regards Xiao
