The file /proc/mdstat will tell you most of what you need to know when you are working with a software RAID array in linux.
cat /proc/mdstat
Returns this on my file server at home
Personalities : [raid6] [raid5] [raid4] [raid1]
md0 : active raid1 sda1[0] sdb1[1] sdc1[2](S) sdd1[3](S) sde1[4](S) sdf1[5](S)
104320 blocks [2/2] [UU]
resync=DELAYED
md1 : active raid5 sdf2[6] sde2[4] sdd2[3] sdc2[2] sdb2[1] sda2[0]
2441396480 blocks level 5, 256k chunk, algorithm 2 [6/5] [UUUUU_]
[========>............] recovery = 43.8% (213871868/488279296) finish=88.4min speed=51686K/sec
unused devices: <none>
The first line of interest is the personalities. This tells you what RAID levels the box supports.
Next is the array name (md0) Whether it’s active or not, it’s RAID level (1) and the devices and partitions that make up this array (sda2-f2). Notice that 4 of the disks have an S behind them to indicate they are spare drives.
The next line is the number of blocks the array has (this can be translated into size) and a section in square brackets. The first number in the brackets is the number of drives that are configured to be active in this array, the next number is the number of drives that actually are active. In my case, 2 are configured and 2 are active, so 2/2. The last box tells us what state those drives are in, U for Up.
Looking at the md1 line, we see a few differences. The immediate difference is the additional information about chunk size and algorithms. this is a RAID 5 set, so that information makes sense, but I’ll let you research it on your own. We can see on this array, 6 drives are configured but only 5 are active. That’s because the array is rebuilding so one of the drives is out of sync. When the rebuild is complete, all of the drive spaces will be U’s. Note that the out of sync drive is flagged as “_” which means the drive isn’t failed, it’s just building.
