Re: init scripts problems ?

Peter T. Breuer wrote:

> 
> Are you saying that the md device wants to talk to nbd AFTER it has
> been set faulty? That's clearly not part of any feasible MD design :-).

It does !

mdadm --fail /dev/mdo /dev/nda :

raid1: Disk failure on nda, disabling device.
         Operation continuing on 1 devices
raid1: made bitmap cbeabe80 at events 0:1e1
md0: md_check_recovery 2: repair bit 0 on sb cf703cc0 preserved

mdadm -S /dev/md0 :

destroying bitmap cbeabe80
bitmap: 4 pages in use
bitmap: 4 pages allocated
bitmap: 0 pages pre-allocated
md: md0 stopped.
md0: notifying dev 2b00000 it is no longer in array
ENBD #4375[1]: enbd_ioctl received BLKMDUNTFY, now out of raid 900000
ENBD #4387[1]: enbd_ioctl cleared show_errs on nda

It occurs AFTER the device is failed !


md: unbind<nda>
md: export_rdev(nda)
md 1243: unlock_rdev nulls bdev
md: WARNING: delaying free of exported rdev cbeaba80
md0: notifying dev 800000 it is no longer in array
md: unbind<sda>
md: export_rdev(sda)
md 1243: unlock_rdev nulls bdev
md: WARNING: delaying free of exported rdev cf5c9840


I will try with stock RAID1 driver...