错误状态：
npc:/cangku# cat /proc/mdstat
Personalities :
md0 : inactive hdc5[0](S) sda5[3](S) hdb5[2](S) hda5[1](S)
      43053568 blocks
       
unused devices: <none>

npc:/cangku# /etc/init.d/mdadm restart
Stopping MD monitoring service: mdadm --monitor.
Starting MD monitoring service: mdadm --monitor.

npc:/cangku# /etc/init.d/mdadm-raid restart
Stopping MD array md0...done (stopped).
Assembling MD array md0...failed (not enough devices).
Generating udev events for MD arrays...done.

恢复md0运行失败：
npc:/cangku# mdadm --manage -R /dev/md0
mdadm: failed to run array /dev/md0: Input/output error

查看md0详细信息， 有2个硬盘离线：
npc:/cangku# mdadm --misc -D /dev/md0
/dev/md0:
        Version : 00.90
  Creation Time : Mon Jan  1 08:12:15 2007
     Raid Level : raid5
  Used Dev Size : 10763392 (10.26 GiB 11.02 GB)
   Raid Devices : 4
  Total Devices : 2
Preferred Minor : 0
    Persistence : Superblock is persistent

    Update Time : Wed Oct 29 19:23:53 2008
          State : active, degraded, Not Started
 Active Devices : 2
Working Devices : 2
 Failed Devices : 0
  Spare Devices : 0

         Layout : left-symmetric
     Chunk Size : 64K

           UUID : 213d6f53:4c19e916:df285044:5b83b7ec
         Events : 0.1481220

    Number   Major   Minor   RaidDevice State
       0      22        5        0      active sync   /dev/hdc5
       1       3        5        1      active sync   /dev/hda5
       2       0        0        2      removed
       3       0        0        3      removed

把2个硬盘添加回去： 或者使用(mdadm -S /dev/md0; mdadm -A dev/md0 /dev/hd[abcd]5)  停止md0,然后重新建立，数据不会丢失。
添加命令:
npc:/cangku# mdadm --manage /dev/md0 -a /dev/sda5
mdadm: re-added /dev/sda5

npc:/cangku# mdadm --manage /dev/md0 -a /dev/hdb5
mdadm: re-added /dev/hdb5

npc:/cangku# cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4]
md0 : inactive hdb5[2] sda5[3] hdc5[0] hda5[1]
      43053568 blocks
      
unused devices: <none>

npc:/cangku# mdadm  -R /dev/md0
mdadm: started /dev/md0



如果磁盘有坏道， 此步骤就会失败， 所以要事先进行磁盘扫描修复:http://anheng.com.cn/news/html/net_admin_blog/linux_badblocks_online_fix.html
完成之后的情况:
npc:/cangku# cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4]
md0 : active raid5 hdb5[2] sda5[3] hdc5[0] hda5[1]
      32290176 blocks level 5, 64k chunk, algorithm 2 [4/4] [UUUU]
     
unused devices: <none>

npc:/cangku# /etc/init.d/lvm2  start
Setting up LVM Volume Groups  Reading all physical volumes.  This may take a while...
  Found volume group "main" using metadata type lvm2
  6 logical volume(s) in volume group "main" now active
.
npc:/cangku# ls /dev/main/
liushiwei  root       swap       tmp        usr        var  
 


如果重建raid，　还是不能解决，此种情况一般会有多于一个硬盘处于spare状态，

Device Role : spare
正常应该是
Device Role : Active device 0

此时就要使用强制方式重建,忽略spare状态:
mdadm --create --assume-clean --level=5 --raid-devices=6 /dev/md0 /dev/sda5 /dev/sdb5 /dev/sdc5 /dev/sdd5 /dev/sde5 /dev/sdf5
参考:https://raid.wiki.kernel.org/index.php/RAID_Recovery