mdadm详解

★mdadm简介

我们可以使用 man mdadm命令来查看mdadm的帮助信息：

[root@localhost mdadm-2.6.2]# man mdadm

关于 man mdadm的详细信息请查看 4楼

☆mdadm用法

基本语法：

mdadm [mode] <raid-device> [options] <component-devices>

目前支持：

LINEAR, RAID0(striping), RAID1(mirroring), RAID4, RAID5, RAID6, RAID10, MULTIPATH和FAULTY

模式(7种)：

• Assemble：加入一个以前定义的阵列
• Build：创建一个没有超级块的阵列
• Create：创建一个新的阵列，每个设备具有超级块
• Manage： 管理阵列(如添加和删除)
• Misc：允许单独对阵列中的某个设备进行操作(如停止阵列)
• Follow or Monitor:监控RAID的状态
• Grow：改变RAID的容量或阵列中的设备数目

选项：
-A, --assemble：加入一个以前定义的阵列
-B, --build：创建一个没有超级块的阵列(Build a legacy array without superblocks.)
-C, --create：创建一个新的阵列
-F, --follow, --monitor：选择监控(Monitor)模式
-G, --grow：改变激活阵列的大小或形态
-I, --incremental：添加一个单独的设备到合适的阵列，并可能启动阵列
--auto-detect：请求内核启动任何自动检测到的阵列
-h, --help：帮助信息，用在以上选项后，则显示该选项信息
--help-options：显示更详细的帮助
-V, --version：打印mdadm的版本信息
-v, --verbose：显示细节
-b, --brief：较少的细节。用于 --detail 和 --examine 选项
-Q, --query：查看一个device，判断它为一个 md device 或是 一个 md 阵列的一部分
-D, --detail：打印一个或多个 md device 的详细信息
-E, --examine：打印 device 上的 md superblock 的内容
-c, --config= ：指定配置文件，缺省为 /etc/mdadm.conf
-s, --scan：扫描配置文件或 /proc/mdstat以搜寻丢失的信息。配置文件/etc/mdadm.conf



★使用mdadm创建RAID5

Create ( mdadm --create)模式用来创建一个新的阵列。 在这里我们首先使用 mdadm --create --help查看一下帮助：

[root@localhost mdadm-2.6.2]# mdadm --create --help
Usage:  mdadm --create device -chunk=X --level=Y --raid-devices=Z devices

This usage will initialise a new md array, associate some
devices with it, and activate the array.   In order to create an
array with some devices missing, use the special word 'missing' in
place of the relevant device name.

Before devices are added, they are checked to see if they already contain
raid superblocks or filesystems.  They are also checked to see if
the variance in device size exceeds 1%.
If any discrepancy is found, the user will be prompted for confirmation
before the array is created.  The presence of a '--run' can override this
caution.

If the --size option is given then only that many kilobytes of each
device is used, no matter how big each device is.
If no --size is given, the apparent size of the smallest drive given
is used for raid level 1 and greater, and the full device is used for
other levels.

Options that are valid with --create (-C) are:
 --bitmap=          : Create a bitmap for the array with the given filename
 --chunk=      -c   : chunk size of kibibytes
 --rounding=        : rounding factor for linear array (==chunk size)
 --level=      -l   : raid level: 0,1,4,5,6,linear,multipath and synonyms
 --parity=     -p   : raid5/6 parity algorithm: {left,right}-{,a}symmetric
 --layout=          : same as --parity
 --raid-devices= -n : number of active devices in array
 --spare-devices= -x: number of spares (eXtras) devices in initial array
 --size=       -z   : Size (in K) of each drive in RAID1/4/5/6/10 - optional
 --force       -f   : Honour devices as listed on command line.  Don't
                    : insert a missing drive for RAID5.
 --run         -R   : insist of running the array even if not all
                    : devices are present or some look odd.
 --readonly    -o   : start the array readonly - not supported yet.
 --name=       -N   : Textual name for array - max 32 characters
 --bitmap-chunk=    : bitmap chunksize in Kilobytes.
 --delay=      -d   : bitmap update delay in seconds.

接下来我们使用 mdadm创建在/dev/md0上创建一个由 sdb、sdc、sdd3块盘组成(另外1块盘 sde为热备)的RAID5：

[root@localhost mdadm-2.6.2]# mdadm --create --verbose /dev/md0 --level=raid5 --raid-devices=3 /dev/sdb /dev/sdc /dev/sdd --spare-devices=1 /dev/sde
mdadm: layout defaults to left-symmetric
mdadm: chunk size defaults to 64K
mdadm: size set to 8388544K
mdadm: array /dev/md0 started.

每个 mdadm的选项都有一个 缩写的形式，例如，上面我们创建RAID 5的命令可以使用下列的缩写形式：

[root@localhost mdadm-2.6.2]# mdadm -Cv /dev/md0 -l5 -n3 /dev/sdb /dev/sdc /dev/sdd -x1 /dev/sde

二者的效果是相同的。

★查看RAID状态

接下来我们使用 cat /proc/mdstat命令来查看一下RAID的状态，我们也可以利用 watch命令来每隔一段时间刷新 /proc/mdstat的输出。使用 CTRL+C可以取消。

[root@localhost mdadm-2.6.2]# watch -n 10 'cat /proc/mdstat'
Every 10s: cat /proc/mdstat                             Thu May 24 11:53:46 2007

Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdd[4] sde[3] sdc[1] sdb[0]
     16777088 blocks level 5, 64k chunk, algorithm 2 [3/2] [UU_]
     [====>................]  recovery = 24.0% (2016364/8388544) finish=10.2min
speed=10324K/sec
unused devices: <none>
[root@localhost mdadm-2.6.2]#

接下来我们为阵列创建文件系统：

[root@localhost mdadm-2.6.2]# mkfs.ext3 /dev/md0
mke2fs 1.34 (25-Jul-2003)
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
2097152 inodes, 4194272 blocks
209713 blocks (5.00%) reserved for the super user
First data block=0
128 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks:
       32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
       4096000

Writing inode tables: done
Creating journal (8192 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 34 mounts or
180 days, whichever comes first.  Use tune2fs -c or -i to override.
You have new mail in /var/spool/mail/root

我们尝试向RAID中写入一个test2文件：

[root@localhost eric4ever]# vi test2
copy succeed!
eric@tlf
[url]http://eric4ever.googlepages.com/[/url]
done!
[root@localhost eric4ever]# ls
LATEST.tgz  mdadm-2.6.2  test2
[root@localhost eric4ever]# mount /dev/md0 /mnt/md0
[root@localhost eric4ever]# df -lh
Filesystem            Size  Used Avail Use% Mounted on
/dev/sda1             2.9G  1.8G  1.1G  63% /
/dev/sda3             4.6G   33M  4.3G   1% /opt
none                  125M     0  125M   0% /dev/shm
/dev/md0               16G   33M   15G   1% /mnt/md0
[root@localhost eric4ever]# ls /mnt/md0
lost+found
[root@localhost eric4ever]# cp ./test2 /mnt/md0
[root@localhost eric4ever]# ls /mnt/md0
lost+found  test2
[root@localhost eric4ever]# ls -lh /mnt/md0
total 20K
drwx------    2 root     root          16K May 24 11:55 lost+found
-rw-r--r--    1 root     root           63 May 24 11:56 test2

使用 mdadm --detail /dev/md0(或 mdadm -D /dev/md0)命令以及 cat /proc/mdstat命令可以查看RAID设备的状态：

[root@localhost eric4ever]# mdadm -D /dev/md0  (或mdadm --detail /dev/md0)
/dev/md0:
       Version : 00.90.00
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
    Array Size : 16777088 (16.00 GiB 17.18 GB)
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0
   Persistence : Superblock is persistent

   Update Time : Thu May 24 13:45:36 2007
         State : active, degraded, recovering
Active Devices : 2
Working Devices : 4
Failed Devices : 1
 Spare Devices : 2

        Layout : left-symmetric
    Chunk Size : 64K

Rebuild Status : 16% complete

          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
        Events : 0.2

   Number   Major   Minor   RaidDevice State
      0       8       16        0      active sync   /dev/sdb
      1       8       32        1      active sync   /dev/sdc
      2       0        0        2      removed

      3       8       64        3      spare   /dev/sde
      4       8       48        4      spare   /dev/sdd

通过 mdadm -D命令，我们可以查看RAID的 版本、创建的时间、RAID级别、阵列容量、可用空间、设备数量、超级块、更新时间、各个设备的状态、RAID算法以及块大小等信息，通过上面的信息我们可以看到目前RAID正处于 重建过程之中，进度为 16%，其中 /dev/sdb和 /dev/sdc两块盘已经 同步。使用 watch命令每个30秒刷新一下查看的进度：

[root@localhost eric4ever]# watch -n 30 'cat /proc/mdstat'
Every 30s: cat /proc/mdstat                             Thu May 24 13:55:56 2007

Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdd[4] sde[3] sdc[1] sdb[0]
     16777088 blocks level 5, 64k chunk, algorithm 2 [3/2] [UU_]
     [==============>......]  recovery = 72.3% (6067444/8388544) finish=3.7min
speed=10324K/sec
unused devices: <none>

当进度进行到 100%时，显示如下：

Every 30s: cat /proc/mdstat                             Thu May 24 14:00:57 2007

Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdd[2] sde[3] sdc[1] sdb[0]
     16777088 blocks level 5, 64k chunk, algorithm 2 [3/3] [UUU]

unused devices: <none>

这是我们再使用 mdadm -D命令查看一下：

[root@localhost eric4ever]# mdadm -D /dev/md0  (或mdadm --detail /dev/md0)
/dev/md0:
       Version : 00.90.00
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
    Array Size : 16777088 (16.00 GiB 17.18 GB)
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0
   Persistence : Superblock is persistent

   Update Time : Thu May 24 13:59:48 2007
         State : active
Active Devices : 3
Working Devices : 4
Failed Devices : 1
 Spare Devices : 1

        Layout : left-symmetric
    Chunk Size : 64K

          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
        Events : 0.3

   Number   Major   Minor   RaidDevice State
      0       8       16        0      active sync   /dev/sdb
      1       8       32        1      active sync   /dev/sdc
      2       8       48        2      active sync   /dev/sdd

      3       8       64        3      spare   /dev/sde

这时我们可以发现 sdb、sdc、sdd三块盘均已经同步。

一般来说，一个新阵列被创建后我们最好创建一个/etc/mdadm.conf文件。没有该文件在激活阵列时我们就得指定更详细的信息，为方便，我们使用下列命令：

[root@localhost eric4ever]# mdadm --detail --scan
ARRAY /dev/md0 level=raid5 num-devices=3 spares=1 UUID=4b15050e:7d0c477d:98ed7d00:0f3c29e4
[root@localhost eric4ever]# mdadm --detail --scan >> /etc/mdadm.conf
[root@localhost eric4ever]# cat /etc/mdadm.conf
ARRAY /dev/md0 level=raid5 num-devices=3 spares=1 UUID=4b15050e:7d0c477d:98ed7d00:0f3c29e4

★启动停止RAID

使用 --stop或 -S命令选项可以停止运行的阵列( 注意： 停止前必须先umount)：

[root@localhost eric4ever]# umount /mnt/md0
[root@localhost eric4ever]# mdadm -S /dev/md0  (或mdadm --stop /dev/md0)
mdadm: stopped /dev/md0

重新启动可以使用：

[root@localhost eric4ever]# mdadm -As /dev/md0
mdadm: /dev/md0 has been started with 3 drives and 1 spare.

★模拟故障

同raidtools一样，mdadm也可以软件模拟故障，命令选项为 --fail或 --set-faulty：

[root@localhost eric4ever]# mdadm --set-faulty --help
Usage: mdadm arraydevice options component devices...

This usage is for managing the component devices within an array.
The --manage option is not needed and is assumed if the first argument
is a device name or a management option.
The first device listed will be taken to be an md array device, and
subsequent devices are (potential) components of that array.

Options that are valid with management mode are:
 --add         -a   : hotadd subsequent devices to the array
 --remove      -r   : remove subsequent devices, which must not be active
 --fail        -f   : mark subsequent devices a faulty
 --set-faulty       : same as --fail
 --run         -R   : start a partially built array
 --stop        -S   : deactivate array, releasing all resources
 --readonly    -o   : mark array as readonly
 --readwrite   -w   : mark array as readwrite
[root@localhost eric4ever]# mdadm --fail --help
Usage: mdadm arraydevice options component devices...

This usage is for managing the component devices within an array.
The --manage option is not needed and is assumed if the first argument
is a device name or a management option.
The first device listed will be taken to be an md array device, and
subsequent devices are (potential) components of that array.

Options that are valid with management mode are:
 --add         -a   : hotadd subsequent devices to the array
 --remove      -r   : remove subsequent devices, which must not be active
 --fail        -f   : mark subsequent devices a faulty
 --set-faulty       : same as --fail
 --run         -R   : start a partially built array
 --stop        -S   : deactivate array, releasing all resources
 --readonly    -o   : mark array as readonly
 --readwrite   -w   : mark array as readwrite

接下来我们模拟 /dev/sdb故障：

[root@localhost eric4ever]# mdadm --manage --set-faulty /dev/md0 /dev/sdb
mdadm: set /dev/sdb faulty in /dev/md0

查看一下系统日志，如果你配置了冗余磁盘，可能会显示如下信息：

   kernel: raid5: Disk failure on sdb, disabling device.
   kernel: md0: resyncing spare disk sde to replace failed disk

检查 /proc/mdstat，如果配置的冗余磁盘可用，阵列可能已经开始重建。

首先我们使用 mdadm --detail /dev/md0命令来查看一下RAID的状态：

[root@localhost eric4ever]# mdadm --detail /dev/md0
/dev/md0:
       Version : 00.90.00
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
    Array Size : 16777088 (16.00 GiB 17.18 GB)
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0
   Persistence : Superblock is persistent

   Update Time : Thu May 24 14:07:55 2007
         State : active, degraded, recovering
Active Devices : 2
Working Devices : 3
Failed Devices : 2
 Spare Devices : 1

        Layout : left-symmetric
    Chunk Size : 64K

Rebuild Status : 3% complete

          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
        Events : 0.6

   Number   Major   Minor   RaidDevice State
      0       8       16        0      faulty spare   /dev/sdb
      1       8       32        1      active sync   /dev/sdc
      2       8       48        2      active sync   /dev/sdd

      3       8       64        3      spare rebuilding   /dev/sde

查看 /proc/mdstat：

[root@localhost eric4ever]# cat /proc/mdstat
Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdb[4] sde[3] sdd[2] sdc[1]
     16777088 blocks level 5, 64k chunk, algorithm 2 [3/2] [_UU]
     [==>..................]  recovery = 10.2% (858824/8388544) finish=12.4min speed=10076K/sec
unused devices: <none>

再查看一下RAID状态：

[root@localhost eric4ever]# mdadm --detail /dev/md0
/dev/md0:
       Version : 00.90.00
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
    Array Size : 16777088 (16.00 GiB 17.18 GB)
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0
   Persistence : Superblock is persistent

   Update Time : Thu May 24 14:08:27 2007
         State : active, degraded, recovering
Active Devices : 2
Working Devices : 4
Failed Devices : 1
 Spare Devices : 2

        Layout : left-symmetric
    Chunk Size : 64K

Rebuild Status : 11% complete

          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
        Events : 0.8

   Number   Major   Minor   RaidDevice State
      0       0        0        0      removed
      1       8       32        1      active sync   /dev/sdc
      2       8       48        2      active sync   /dev/sdd

      3       8       64        3      spare   /dev/sde
      4       8       16        4      spare   /dev/sdb

已经完成到 11%了。查看一下日志消息：

[root@localhost eric4ever]# tail /var/log/messages
May 24 14:08:27 localhost kernel:  --- rd:3 wd:2 fd:1
May 24 14:08:27 localhost kernel:  disk 0, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
May 24 14:08:27 localhost kernel:  disk 1, s:0, o:1, n:1 rd:1 us:1 dev:sdc
May 24 14:08:27 localhost kernel:  disk 2, s:0, o:1, n:2 rd:2 us:1 dev:sdd
May 24 14:08:27 localhost kernel: RAID5 conf printout:
May 24 14:08:27 localhost kernel:  --- rd:3 wd:2 fd:1
May 24 14:08:27 localhost kernel:  disk 0, s:0, o:0, n:0 rd:0 us:0 dev:[dev 00:00]
May 24 14:08:27 localhost kernel:  disk 1, s:0, o:1, n:1 rd:1 us:1 dev:sdc
May 24 14:08:27 localhost kernel:  disk 2, s:0, o:1, n:2 rd:2 us:1 dev:sdd
May 24 14:08:27 localhost kernel: md: cannot remove active disk sde from md0 ...

使用 mdadm -E命令查看一下 /dev/sdb的情况：

[root@localhost eric4ever]# mdadm -E /dev/sdb
/dev/sdb:
         Magic : a92b4efc
       Version : 00.90.00
          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
    Array Size : 16777088 (16.00 GiB 17.18 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0

   Update Time : Thu May 24 14:08:27 2007
         State : active
Active Devices : 2
Working Devices : 4
Failed Devices : 1
 Spare Devices : 2
      Checksum : a6a19662 - correct
        Events : 0.8

        Layout : left-symmetric
    Chunk Size : 64K

     Number   Major   Minor   RaidDevice State
this     4       8       16        4      spare   /dev/sdb

  0     0       0        0        0      faulty removed
  1     1       8       32        1      active sync   /dev/sdc
  2     2       8       48        2      active sync   /dev/sdd
  3     3       8       64        3      spare   /dev/sde
  4     4       8       16        4      spare   /dev/sdb

自动修复完成后，我们再查看一下RAID的状态：

[root@localhost eric4ever]# mdadm --detail /dev/md0
/dev/md0:
       Version : 00.90.00
 Creation Time : Thu May 24 13:45:35 2007
    Raid Level : raid5
    Array Size : 16777088 (16.00 GiB 17.18 GB)
 Used Dev Size : 8388544 (8.00 GiB 8.59 GB)
  Raid Devices : 3
 Total Devices : 5
Preferred Minor : 0
   Persistence : Superblock is persistent

   Update Time : Thu May 24 14:21:54 2007
         State : active
Active Devices : 3
Working Devices : 4
Failed Devices : 1
 Spare Devices : 1

        Layout : left-symmetric
    Chunk Size : 64K

          UUID : 4b15050e:7d0c477d:98ed7d00:0f3c29e4
        Events : 0.9

   Number   Major   Minor   RaidDevice State
      0       8       64        0      active sync   /dev/sde
      1       8       32        1      active sync   /dev/sdc
      2       8       48        2      active sync   /dev/sdd

      4       8       16        4      spare   /dev/sdb
[root@localhost eric4ever]# cat /proc/mdstat
Personalities : [raid5]
read_ahead 1024 sectors
md0 : active raid5 sdb[4] sde[0] sdd[2] sdc[1]
     16777088 blocks level 5, 64k chunk, algorithm 2 [3/3] [UUU]

unused devices: <none>

我们可以看到 /dev/sde已经替换了 /dev/sdb。看看系统的日志消息：

[root@localhost eric4ever]# tail /var/log/messages
May 24 14:21:54 localhost kernel:  --- rd:3 wd:3 fd:0
May 24 14:21:54 localhost kernel:  disk 0, s:0, o:1, n:0 rd:0 us:1 dev:sde
May 24 14:21:54 localhost kernel:  disk 1, s:0, o:1, n:1 rd:1 us:1 dev:sdc
May 24 14:21:54 localhost kernel:  disk 2, s:0, o:1, n:2 rd:2 us:1 dev:sdd
May 24 14:21:54 localhost kernel: md: updating md0 RAID superblock on device
May 24 14:21:54 localhost kernel: md: sdb [events: 00000009]<6>(write) sdb's sb offset: 8388544
May 24 14:21:54 localhost kernel: md: sde [events: 00000009]<6>(write) sde's sb offset: 8388544
May 24 14:21:54 localhost kernel: md: sdd [events: 00000009]<6>(write) sdd's sb offset: 8388544
May 24 14:21:54 localhost kernel: md: sdc [events: 00000009]<6>(write) sdc's sb offset: 8388544
May 24 14:21:54 localhost kernel: md: recovery thread got woken up ...

recovery thread got woken up ...

这时我们可以从 /dev/md0中移除 /dev/sdb设备：

[root@localhost eric4ever]# mdadm /dev/md0 -r /dev/sdb
mdadm: hot removed /dev/sdb

类似地，我们可以使用下列命令向 /dev/md0中添加一个设备：

[root@localhost eric4ever]# mdadm /dev/md0 --add /dev/sdf

★监控RAID

mdadm的监控模式提供一些实用的功能，你可以使用下列命令来监控 /dev/md0，delay参数意味着检测的时间间隔，这样紧急事件和严重的错误会及时发送给系统管理员：

[root@localhost eric4ever]# mdadm --monitor --mail=eric4ever@localhost --delay=300 /dev/md0

当使用监控模式时，mdadm不会退出，你可以使用下列命令：

[root@localhost eric4ever]# nohup mdadm --monitor --mail=eric4ever@localhost --delay=300 /dev/md0 &
[1] 3113
[root@localhost eric4ever]# nohup: appending output to `nohup.out'

Trackback: http://tb.blog.csdn.net/TrackBack.aspx?PostId=1828868