RAID演习记
背景
- Buffalo LS-WVL
- RAID1
- 两块盘
- 各1TB
检查
root@ubuntu:~/mnt# mdadm -D /dev/md2
/dev/md2:
Version : 1.2
Creation Time : Fri Dec 14 14:36:06 2012
Raid Level : raid1
Array Size : 961748840 (917.20 GiB 984.83 GB)
Used Dev Size : 961748840 (917.20 GiB 984.83 GB)
Raid Devices : 2
Total Devices : 1
Persistence : Superblock is persistent
Update Time : Mon Nov 28 21:41:47 2016
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 0
Spare Devices : 0
Name : LS-WVL9BE:2
UUID : 94d65994:7f391a0e:b959e3cc:407963fb
Events : 6617621
Number Major Minor RaidDevice State
0 0 0 0 removed
1 8 38 1 active sync /dev/sdc6
root@ubuntu:~/mnt# mdadm --examine /dev/sdc5
/dev/sdc5:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : f245bd1b:956cf823:2c2f8665:629c2eec
Name : LS-WVL-EM9BE:10
Creation Time : Fri Dec 14 14:20:07 2012
Raid Level : raid1
Raid Devices : 2
Avail Dev Size : 2000872 (977.15 MiB 1024.45 MB)
Array Size : 1000436 (977.15 MiB 1024.45 MB)
Data Offset : 24 sectors
Super Offset : 8 sectors
State : clean
Device UUID : f5f2f22d:b1144de0:9a190acb:5c808f34
Update Time : Mon Nov 28 21:11:40 2016
Checksum : d2c31913 - correct
Events : 40287
Device Role : Active device 1
Array State : AA ('A' == active, '.' == missing)
root@ubuntu:~/mnt# cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md2 : active raid1 sdc6[1]
961748840 blocks super 1.2 [2/1] [_U]
md10 : active raid1 sdd5[0] sdc5[1]
1000436 blocks super 1.2 [2/2] [UU]
md1 : active raid1 sdc2[1]
4999156 blocks super 1.2 [2/1] [_U]
md0 : active raid1 sdc1[1]
1000384 blocks [2/1] [_U]
unused devices: <none>
root@ubuntu:~# mdadm --detail --scan
ARRAY /dev/md0 metadata=0.90 spares=1 UUID=6664ed44:3a6468f0:fc4e250d:63bd81ed
ARRAY /dev/md/1 metadata=1.2 spares=1 name=LS-WVL-EM9BE:1 UUID=915ffb58:2777fdc1:e33a57b8:b0a1cd76
ARRAY /dev/md/10 metadata=1.2 name=LS-WVL-EM9BE:10 UUID=f245bd1b:956cf823:2c2f8665:629c2eec
ARRAY /dev/md/2 metadata=1.2 spares=1 name=LS-WVL9BE:2 UUID=94d65994:7f391a0e:b959e3cc:407963f
正常情况: 插两块盘挂载
mdadm --assemble --scan
mkdir mnt/{0,1,10,2}
mount /dev/md0 mnt/0/
mount /dev/md1 mnt/1/
mount /dev/md2 mnt/2/
mount /dev/md10 mnt/10/ #挂载不上去
异常情况: 两个RAID1, 只插一块盘, 挂载
mdadm --stop /dev/md{0,1,10,2}
mdadm --assemble --scan
mount -o rw /dev/md2 mnt/2/
再插回来两块盘
由于上一步往里写了一些数据, 两块盘插回来状态已经是degraded了.
root@ubuntu:~/mnt# mdadm -A --scan
mdadm: /dev/md0 is already in use.
mdadm: /dev/md/1 is already in use.
mdadm: /dev/md/2 has been started with 1 drive (out of 2).
mdadm: /dev/md0 is already in use.
mdadm: /dev/md/1 is already in use.
root@ubuntu:~/mnt# mdadm -D /dev/md2
/dev/md2:
Version : 1.2
Creation Time : Fri Dec 14 14:36:06 2012
Raid Level : raid1
Array Size : 961748840 (917.20 GiB 984.83 GB)
Used Dev Size : 961748840 (917.20 GiB 984.83 GB)
Raid Devices : 2
Total Devices : 1
Persistence : Superblock is persistent
Update Time : Mon Nov 28 21:41:47 2016
State : clean, degraded
Active Devices : 1
Working Devices : 1
Failed Devices : 0
Spare Devices : 0
Name : LS-WVL9BE:2
UUID : 94d65994:7f391a0e:b959e3cc:407963fb
Events : 6617621
Number Major Minor RaidDevice State
0 0 0 0 removed <---这里
1 8 38 1 active sync /dev/sdc6
root@ubuntu:~/mnt# mdadm --manage /dev/md2 -a /dev/sdd6
mdadm: added /dev/sdd6
root@ubuntu:~/mnt# mdadm -D /dev/md2
/dev/md2:
Version : 1.2
Creation Time : Fri Dec 14 14:36:06 2012
Raid Level : raid1
Array Size : 961748840 (917.20 GiB 984.83 GB)
Used Dev Size : 961748840 (917.20 GiB 984.83 GB)
Raid Devices : 2
Total Devices : 2
Persistence : Superblock is persistent
Update Time : Mon Nov 28 22:03:55 2016
State : clean, degraded, recovering
Active Devices : 1
Working Devices : 2
Failed Devices : 0
Spare Devices : 1
Rebuild Status : 0% complete
Name : LS-WVL9BE:2
UUID : 94d65994:7f391a0e:b959e3cc:407963fb
Events : 6617623
Number Major Minor RaidDevice State
2 8 54 0 spare rebuilding /dev/sdd6
1 8 38 1 active sync /dev/sdc6
估计要重建很久.
看一下速度:
root@ubuntu:~/mnt# cat /proc/mdstat
Personalities : [linear] [multipath] [raid0] [raid1] [raid6] [raid5] [raid4] [raid10]
md2 : active raid1 sdd6[2] sdc6[1]
961748840 blocks super 1.2 [2/1] [_U]
[>....................] recovery = 0.5% (4856512/961748840) finish=1132.9min speed=14076K/sec
md10 : active raid1 sdd5[0] sdc5[1]
1000436 blocks super 1.2 [2/2] [UU]
md1 : active raid1 sdc2[1]
4999156 blocks super 1.2 [2/1] [_U]
md0 : active raid1 sdc1[1]
1000384 blocks [2/1] [_U]
unused devices: <none>
此部分参考: https://www.thomas-krenn.com/en/wiki/Mdadm_recover_degraded_Array
番外篇: RAID5 演练
留作习题. 以下给出一些提示.
观察所有的状态
watch lsblk
watch cat /proc/mdstat
watch sudo mdadm -D /dev/md0
mdadm --create --verbose /dev/md0 --level=5 --raid-devices=3 /dev/sdb1 /dev/sdc1 /dev/sdd1 --spare-devices=1 /dev/sde1
mdadm --assemble --scan
# 也可以加个 --no-degrade开关
mdadm -D --scan >> /etc/mdadm/mdadm.conf #先备份一下组合的UUID什么的
比如RAID5, 一共3个U盘, 插了2个, 可以工作在degrade模式, 照常读写. 插上第3个之后, 要重新
mdadm --manage --add /dev/md0 /dev/sdX
RAID5, 3个, 再加一个, 第4个可以成为热备, 拔掉其中一个, 过不久就会自动进行spare rebuild
还可以把某一个用mdadm --manage --replace /dev/md0 /dev/sdX标记为replace, 这样就会先去在热备的盘上进行rebuild, 完了之后就把标记为replace的盘, 标记成failed. 接着再把它remove就行.
详见:
mdadm --manage --help
反正感觉上就是那个mdadm的daemon得一直活着. 如果要把整个阵列拿下去的话, 要
mdadm --stop /dev/md0
再物理拔掉.
找几个U盘试一下就踏实多了… 以前一直小心翼翼的不敢动
PS: 我是拿4G/4G/16G组RAID5, 阵列可用8G. 试图向里面加一个2G的盘, 提示容量过小不可添加.
See Also
- macOS上使用RAID
diskutil listdiskutil AppleRAID create stripe MyBigNewRaid JHFS+ disk4 disk5- 第三方的SoftRAID操作更友好一些, 可惜收費
- Kernel.org的Wiki