]> git.proxmox.com Git - mirror_ubuntu-kernels.git/commit
DM RAID: Add ability to restore transiently failed devices on resume
authorJonathan Brassow <jbrassow@redhat.com>
Thu, 2 May 2013 19:19:24 +0000 (14:19 -0500)
committerNeilBrown <neilb@suse.de>
Thu, 13 Jun 2013 22:10:24 +0000 (08:10 +1000)
commit9092c02d943515b3c9ffd5d0003527f8cc1dd77b
tree686180c0788704a8cbde2080b83e7ace1255b277
parent25e33ed9c711c8d64c403a17d4a2cdeac213800b
DM RAID: Add ability to restore transiently failed devices on resume

DM RAID: Add ability to restore transiently failed devices on resume

This patch adds code to the resume function to check over the devices
in the RAID array.  If any are found to be marked as failed and their
superblocks can be read, an attempt is made to reintegrate them into
the array.  This allows the user to refresh the array with a simple
suspend and resume of the array - rather than having to load a
completely new table, allocate and initialize all the structures and
throw away the old instantiation.

Signed-off-by: Jonathan Brassow <jbrassow@redhat.com>
Signed-off-by: NeilBrown <neilb@suse.de>
Documentation/device-mapper/dm-raid.txt
drivers/md/dm-raid.c
drivers/md/raid1.c
drivers/md/raid10.c