raid disk not running

MartenH · 08-12-2010, 10:30 AM

Hello,

I just configured two raid setups but after a reboot they are not mounted and seem to be inactive.

md127 = sde1, sdf1 and sdi1 (raid 5)
md0 = sda1 and sdh1 (raid 0)

Code:

root@server /]# cat /proc/mdstat 
Personalities : 
md127 : inactive sdf1[1](S) sde1[2](S)
      78156032 blocks
       
md0 : inactive sda1[0](S)
      488382977 blocks super 1.2
       
unused devices: <none>

Code:

[root@server /]# fdisk -l | grep "Disk /"
Disk /dev/sda: 500.1 GB, 500107862016 bytes
Disk /dev/sdb: 80.0 GB, 80026361856 bytes
Disk /dev/sdc: 122.9 GB, 122942324736 bytes
Disk /dev/sdd: 160.0 GB, 160041885696 bytes
Disk /dev/sde: 40.0 GB, 40020664320 bytes
Disk /dev/sdf: 40.0 GB, 40020664320 bytes
Disk /dev/sdg: 1500.3 GB, 1500301910016 bytes
Disk /dev/sdh: 251.0 GB, 251000193024 bytes
Disk /dev/sdi: 40.0 GB, 40020664320 bytes
Disk /dev/sdj: 500.1 GB, 500107862016 bytes

Code:

[root@server /]# cat /etc/mdadm.conf 
DEVICE /dev/sdi1 /dev/sdf1 /dev/sde1 /dev/sda1 /dev/sdh1
ARRAY /dev/md127 UUID=5dc0cf7a:8c715104:04894333:532a878b auto=yes
ARRAY /dev/md0 UUID=65c49170:733df717:435e470b:3334ee94 auto=yes

As you can see they now show up as inactive. And for some reason sdi1 and sdh1 are not even listed. What can I do to get them back?

To make matters worse I placed some important data on them, and even if I was clever enough to keep an extra copy on another drive, guess which drive that was? So, I need to get them activated as is (at least so I can get the data of them) before I can rebuild them from scratch.

Any assistance is much appreciated!

PS. I'm running Mandriva 2010.1 and rated tehm using the built in disk partitioner.

mostlyharmless · 08-13-2010, 12:54 PM

No doubt a dumb question, but have you tried assembling and mounting them manually at the command line with mdadm, something like mdadm -assemble -scan or even mdadm -assemble {manually add raid parameters and devices}

If not, you could try the posting the output of

mdadm -Ebsc partitions

and we can try to figure it out

MartenH · 08-14-2010, 12:07 PM

Hello again,

After your suggestion I dared running some mdadm commands and got it working. I had to do mdadm --stop first before assembling them with --assemble -v.

The only remaining problem is that I have to do this manually each time the computer is restarted. Is there any way to have them assembled automatically?

jay73 · 08-14-2010, 12:54 PM

You may need to edit your mdadm.conf file. Something like this:

Quote:

# definitions of existing MD arrays
ARRAY /dev/md0 level=raid0 num-devices=2 UUID=b42490f3:ada38e64:0c1c9294:00a2349e
ARRAY /dev/md1 level=raid0 num-devices=2 UUID=3d59e000:18465b7f:052fff41:14d859df

MartenH · 08-14-2010, 01:09 PM

My mdadm.conf says:

Code:

DEVICE /dev/sdi1 /dev/sdf1 /dev/sde1 /dev/sda1 /dev/sdh1
ARRAY /dev/md127 UUID=[numbers] auto=yes
ARRAY /dev/md0 UUID=[numbers] auto=yes

But it still doesn't appear during startup. I am forced to running the -assemble manually.

jay73 · 08-14-2010, 01:12 PM

From the mdadm.conf man page:

Quote:

auto= This option declares to mdadm that it should try to create the device file of the array if it doesn't already exist, or exists but with the wrong device number.

So it says "try", no guarantee that it will succeed. As I suggested, try being more explicit by listing the details (number of devices, spares, raid level). See man mdadm.conf.

MartenH · 08-14-2010, 01:25 PM

Quote:

Originally Posted by jay73

From the mdadm.conf man page:

So it says "try", no guarantee that it will succeed. As I suggested, try being more explicit by listing the details (number of devices, spares, raid level). See man mdadm.conf.

Thanks, I'll give it go.

MartenH · 08-14-2010, 02:25 PM

my mdadm.conf now says:

Code:

DEVICE /dev/sdi1 /dev/sdf1 /dev/sde1 /dev/sda1 /dev/sdh1                                                  
ARRAY /dev/md127 num-devices=3 devices=/dev/sde1,/dev/sdf1,/dev/sdi1 level=5 UUID=5dc0cf7a:8c715104:04894333:532a878b auto=yes
ARRAY /dev/md0 num-devices=2 devices=/dev/sda1,/dev/sdh1 level=0 UUID=65c49170:733df717:435e470b:3334ee94 auto=yes

But it still doesn't work on startup.

mdadm --assemble -scan says it only finds one drive for each set.
[root@hserver /]# mdadm --assemble -scan

Code:

mdadm: /dev/md/127_0 assembled from 1 drive - not enough to start the array.
mdadm: /dev/md/holisrv3:0 assembled from 1 drive - not enough to start the array.
mdadm: No arrays found in config file or automatically

It only finds one device, and it claims there are no arrays in config file, even if they are? And what is with the stange md names?
(unless /etc/mdadm.conf is the wrong config file?)

It feels like there are some conflicting configurations somehow?

MartenH · 08-14-2010, 02:28 PM

Adding additional output as requested previously.

[root@server /]# mdadm -Ebsc partitions
ARRAY /dev/md/0 metadata=1.2 UUID=65c49170:733df717:435e470b:3334ee94 name=holisrv3:0
ARRAY /dev/md127 UUID=5dc0cf7a:8c715104:04894333:532a878b

This sure looks like another config file? But...
[root@server /]# locate mdadm.conf
/etc/mdadm.conf
/usr/share/doc/mdadm/mdadm.conf-example
/usr/share/man/man5/mdadm.conf.5.lzma

Btw. "holisrv3" is the name of the computer.

MartenH · 08-14-2010, 02:44 PM

I noticed my error of stating level=5 instead of level=raid5 and have corrected it. Have not rebooted yet.

Perhaps I should give up on salvaging the current situation, simply erase all traces of the raid sets and do them all over again but manually instead of using the guide?

What say you?

(since I could assemble them manually I have removed all my important data from them).

jay73 · 08-14-2010, 03:13 PM

If you create them manually, you are also responsible for adding entries to the mdadm.conf file so it would be useful to check whether what you have now works or not, if only to give you an idea. Re-creating the raid devices without knowing exactly what the entries should be like would be pointless.

MartenH · 08-14-2010, 03:39 PM

I have now rebooted after correcting the error with level=5. Still same result.

jay73 · 08-14-2010, 04:14 PM

Run mdadm --detail --scan and compare with what you put into mdadm.conf.

MartenH · 08-14-2010, 07:13 PM

Ok.

I removed all sets and rebuilt everything from scratch using this guide. Still no luck.

Current details:
fdisk -l | grep "Disk /" (The disks are there)

Quote:

Disk /dev/sda: 500.1 GB, 500107862016 bytes
Disk /dev/sdb: 80.0 GB, 80026361856 bytes
Disk /dev/sdc: 122.9 GB, 122942324736 bytes
Disk /dev/sdd: 160.0 GB, 160041885696 bytes
Disk /dev/sde: 40.0 GB, 40020664320 bytes
Disk /dev/sdf: 40.0 GB, 40020664320 bytes
Disk /dev/sdg: 1500.3 GB, 1500301910016 bytes
Disk /dev/sdh: 251.0 GB, 251000193024 bytes
Disk /dev/sdi: 40.0 GB, 40020664320 bytes
Disk /dev/sdj: 500.1 GB, 500107862016 bytes

cat /proc/mdstat (why are sdi1 and sdh1 missing?)

Quote:

Personalities :
md0 : inactive sdf1[1](S) sde1[0](S)
78154114 blocks super 1.2

md1 : inactive sda1[0](S)
488382977 blocks super 1.2

unused devices: <none>

/etc/mdadm.conf (Should I add a DEVICE line?)

Quote:

ARRAY /dev/md0 level=raid5 num-devices=3 metadata=1.2 name=holisrv2:0 UUID=935f6f43:780a81ff:e0a5f47f:9a46d24d
devices=/dev/sde1,/dev/sdf1,/dev/sdi1
ARRAY /dev/md1 level=raid0 num-devices=2 metadata=1.2 name=holisrv2:1 UUID=c381c719:983bc278:a033decc:30f6d7bd
devices=/dev/sda1,/dev/sdh1

mdadm --detail --scan

Quote:

mdadm: md device /dev/md1 does not appear to be active.
mdadm: md device /dev/md0 does not appear to be active.

mdadm --assemble --scan

Quote:

mdadm: /dev/md0 is already in use.
mdadm: /dev/md1 is already in use.

mdadm -Ebsc partitions

Quote:

ARRAY /dev/md/1 metadata=1.2 UUID=c381c719:983bc278:a033decc:30f6d7bd name=holisrv2:1
ARRAY /dev/md/0 metadata=1.2 UUID=935f6f43:780a81ff:e0a5f47f:9a46d24d name=holisrv2:0

I then run:

Code:

# mdadm --stop /dev/md0
mdadm: stopped /dev/md0
# mdadm --stop /dev/md1
mdadm: stopped /dev/md1

# mdadm --assemble -v /dev/md0 /dev/sde1 /dev/sdf1 /dev/sdi1
mdadm: looking for devices for /dev/md0
mdadm: /dev/sde1 is identified as a member of /dev/md0, slot 0.
mdadm: /dev/sdf1 is identified as a member of /dev/md0, slot 1.
mdadm: /dev/sdi1 is identified as a member of /dev/md0, slot 2.
mdadm: added /dev/sdf1 to /dev/md0 as 1
mdadm: added /dev/sdi1 to /dev/md0 as 2
mdadm: added /dev/sde1 to /dev/md0 as 0
mdadm: /dev/md0 has been started with 3 drives.

# mdadm --assemble -v /dev/md1 /dev/sda1 /dev/sdh1
mdadm: looking for devices for /dev/md1
mdadm: /dev/sda1 is identified as a member of /dev/md1, slot 0.
mdadm: /dev/sdh1 is identified as a member of /dev/md1, slot 1.
mdadm: added /dev/sdh1 to /dev/md1 as 1
mdadm: added /dev/sda1 to /dev/md1 as 0
mdadm: /dev/md1 has been started with 2 drives.

At this point I can mount them and use them if I want to.

After assembling manually I get the following output:

cat /proc/mdstat

Quote:

Personalities : [raid6] [raid5] [raid4] [raid0]
md1 : active raid0 sda1[0] sdh1[1]
733493248 blocks super 1.2 512k chunks

md0 : active raid5 sde1[0] sdi1[3] sdf1[1]
78153728 blocks super 1.2 level 5, 512k chunk, algorithm 2 [3/3] [UUU]

unused devices: <none>

mdadm --detail --scan

Quote:

ARRAY /dev/md0 metadata=1.2 name=holisrv2:0 UUID=935f6f43:780a81ff:e0a5f47f:9a46d24d
ARRAY /dev/md1 metadata=1.2 name=holisrv2:1 UUID=c381c719:983bc278:a033decc:30f6d7bd

mdadm -Ebsc partitions

Quote:

ARRAY /dev/md/1 metadata=1.2 UUID=c381c719:983bc278:a033decc:30f6d7bd name=holisrv2:1
ARRAY /dev/md/0 metadata=1.2 UUID=935f6f43:780a81ff:e0a5f47f:9a46d24d name=holisrv2:0

So the question remain, why do they refuse to be assembled on startup? I can obviously assemble them manually without any problem.

PS. I noticed that the mdadm service is not started automatically. But starting it with 'service mdadm start' did no difference.

jay73 · 08-14-2010, 11:59 PM

I have used that same how-to without any problem so I am beginning to wonder whether you are not affected by a bug. Maybe it is time for a little detour. I would check with another distro.