linux raid fails to add a hard drive to the raid on boot up

evansmb · 09-22-2014, 12:45 PM

I am setting up a linux raid. It consists of two drives. When the system is rebooted only one drive joins the array. So before the reboot mdstat would look similar to the following:
Personalities : [raid1]
md0 : active raid1 sdc1[2] sdb1[0]
3906876480 blocks super 1.2 [2/1] [U_]
[>....................] recovery = 0.0% (208768/3906876480) finish=1871.1min speed=34794K/sec

How ever after a reboot it mdstat will show something like:

Personalities : [raid1]
md0 : active raid1 sdb1[0]
3906876480 blocks super 1.2 [2/1] [U_]

unused devices: <none>

I then have to add the second drive back to the array manually. Would anyone have an idea of why it will not add the second drive on boot?

smallpond · 09-23-2014, 09:34 AM

You need to have raid built-in to your kernel or loaded by initrd to have it assemble the boot drive. Otherwise it has to mount the drive before it has the module. Be sure to install grub on both disks.

Code:

CONFIG_MD=y
CONFIG_MD_RAID1=y
CONFIG_MD_AUTODETECT=y

evansmb · 09-23-2014, 12:28 PM

Looks like is getting loaded.
[root@netflow 2.6.32-431.17.1.el6.x86_64]# cat /proc/modules
dm_mirror 14384 0 - Live 0xffffffffa002d000
dm_region_hash 12085 1 dm_mirror, Live 0xffffffffa0026000
dm_log 9930 2 dm_mirror,dm_region_hash, Live 0xffffffffa001f000
dm_mod 84337 23 dm_mirror,dm_log, Live 0xffffffffa0000000

looking in the messages logs i see:

Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] 976754645 4096-byte logical blocks: (4.00 TB/3.63 TiB)
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] Write Protect is off
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] Assuming drive cache: write through
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] 976754645 4096-byte logical blocks: (4.00 TB/3.63 TiB)
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] Assuming drive cache: write through
Sep 23 12:25:09 netflow kernel: sdc: sdb1
Sep 23 12:25:09 netflow kernel: sdc1
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] 976754645 4096-byte logical blocks: (4.00 TB/3.63 TiB)
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] Assuming drive cache: write through
Sep 23 12:25:09 netflow kernel: sd 8:0:0:0: [sdc] Attached SCSI disk
Sep 23 12:25:09 netflow kernel: md: bind<sdc>
Sep 23 12:25:09 netflow kernel: md: kicking non-fresh sdc from array!
Sep 23 12:25:09 netflow kernel: md: unbind<sdc>
Sep 23 12:25:09 netflow kernel: md: export_rdev(sdc)

smallpond · 09-24-2014, 09:31 AM

Something is messed up. Your RAID should be on partitions, like sdc1 sdb1. The whole drive sdc should not have a raid superblock. You can check with

Code:

mdadm --examine /dev/sdc