Re[3]: Issue removing failed drive and re adding on raid 6
From: Mikael Abrahamsson <hidden>
Date: 2015-07-04 06:58:13
On Sat, 4 Jul 2015, Justin Stephenson wrote:
quoted hunk ↗ jump to hunk
md/raid:md0: not clean -- starting background reconstruction md/raid:md0: device sdh1 operational as raid disk 5 md/raid:md0: device sdg1 operational as raid disk 4 md/raid:md0: device sdd1 operational as raid disk 1 md/raid:md0: device sde1 operational as raid disk 2 md/raid:md0: device sdb1 operational as raid disk 6 md/raid:md0: device sdc1 operational as raid disk 0 md/raid:md0: allocated 7470kB md/raid:md0: cannot start dirty degraded array. RAID conf printout:--- level:6 rd:7 wd:6disk 0, o:1, dev:sdc1 disk 1, o:1, dev:sdd1 disk 2, o:1, dev:sde1 disk 4, o:1, dev:sdg1 disk 5, o:1, dev:sdh1 disk 6, o:1, dev:sdb1 md/raid:md0: failed to run raid set. md: pers->run() failed ...
This is the interesting part. I don't know why this happens, especially since it didn't help with --assemble --force (which you said before that you did). I found this: http://www.devinzuczek.com/anything-at-all/raid5-cannot-start-dirty-degraded-array-for-mdn/ It indicates that you should be able to do "mdadm --manage /dev/md0 --run". I have never ended up in this situation so I don't know if it'll work.
***** Parent of /run/mdadm does not exist. Maybe set different RUN_DIR= ***** e.g. make RUN_DIR=/dev/.mdadm ***** or set CHECK_RUN_DIR=0 make: *** [check_rundir] Error 1 [justin@BigBlue mdad]$
You don't have to do "make install", you can just do "make" and find the resulting mdadm binary and use ./mdadm to run it from current dir. You probably only need it to actually get the array up and running again, then your system mdadm will be fine. Btw, post "mdadm --version" and "uname -a" so we know what kernel and mdadm version you're running. -- Mikael Abrahamsson email: swmike@swm.pp.se