View previous topic :: View next topic |
Author |
Message |
pascuol n00b
Joined: 03 Dec 2006 Posts: 39
|
Posted: Fri Jul 20, 2012 2:24 pm Post subject: [Solved]Raid doesn't want to extend |
|
|
Hi,
I've a little problem with my RAID
I've got a nice raid.
I wanted to extend the size of my raid like each time I add an bigger hard drive but today it doesnt want..
I don't see wah't the cause is, it never happened before.
So I wanted to know why mdadm doesn't want to get the new sizze of my RAID6
Code: | mdadm -G -z max /dev/md0
mdadm: component size of /dev/md0 has been set to 566516736K
|
instead it should be
Code: | LV Size 1006.03 GiB
|
so, usualy, I putted as failled disk one by one and rebuilted it after changing the size of the LVM, but now it tooks 2 days per disk and it seems unneccessary. Now I've just added the size on each LVM and it should be good enough.
tech specs :
Don't ask why I did it like this .. I've got my own reason:)
I've got an LVM behind my RAID6 with 4 logical disks:
Code: | lvdisplay
--- Logical volume ---
LV Name /dev/RAID/RaidVol1
VG Name RAID
LV UUID hSSEva-AW5H-dVr2-bESF-LpuB-NMSc-sPfwlp
LV Write Access read/write
LV Status available
# open 1
LV Size 1006.03 GiB
Current LE 257543
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:0
--- Logical volume ---
LV Name /dev/RAID/RaidVol2
VG Name RAID
LV UUID PJ8ckn-2j6p-cC2X-CH4e-RQqR-9j6j-VZNzI7
LV Write Access read/write
LV Status available
# open 1
LV Size 1006.03 GiB
Current LE 257543
Segments 2
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:1
--- Logical volume ---
LV Name /dev/RAID/RaidVol3
VG Name RAID
LV UUID IC2XLv-s00R-8eoC-shos-TBSb-BoRZ-xr3biS
LV Write Access read/write
LV Status available
# open 1
LV Size 1006.03 GiB
Current LE 257543
Segments 2
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:2
--- Logical volume ---
LV Name /dev/RAID/RaidVol4
VG Name RAID
LV UUID kWp0Go-xDI4-f7Ve-Fe9U-PdK5-3MCi-NtGJnh
LV Write Access read/write
LV Status available
# open 1
LV Size 1006.03 GiB
Current LE 257543
Segments 2
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:3
|
but mdadm doesn't want to know anything :
Code: | # mdadm -G -z max /dev/md0
mdadm: component size of /dev/md0 has been set to 566516736K
|
It should be the double but ...
any idea ?
Last edited by pascuol on Sun Jul 22, 2012 11:22 am; edited 1 time in total |
|
Back to top |
|
|
frostschutz Advocate
Joined: 22 Feb 2005 Posts: 2977 Location: Germany
|
Posted: Fri Jul 20, 2012 10:51 pm Post subject: |
|
|
/proc/mdstat?
blockdev --getsize64 /dev/RAID/RaidVol*? |
|
Back to top |
|
|
pascuol n00b
Joined: 03 Dec 2006 Posts: 39
|
Posted: Sat Jul 21, 2012 12:22 pm Post subject: |
|
|
frostschutz wrote: | /proc/mdstat?
blockdev --getsize64 /dev/RAID/RaidVol*? |
Code: | cat /proc/mdstat Sat Jul 21 14:21:28 2012
Personalities : [raid6] [raid5] [raid4]
md0 : active raid6 dm-0[4] dm-3[7] dm-2[6] dm-1[5]
1133033472 blocks super 1.2 level 6, 512k chunk, algorithm 2 [4/4] [UUUU]
unused devices: <none>
|
Code: | # blockdev --getsize64 /dev/RAID/RaidVol*
1080213635072
1080213635072
1080213635072
1080213635072
|
|
|
Back to top |
|
|
frostschutz Advocate
Joined: 22 Feb 2005 Posts: 2977 Location: Germany
|
Posted: Sat Jul 21, 2012 2:52 pm Post subject: |
|
|
Hmmm, interesting. Not sure what the problem is.
mdadm -E /dev/RAID/RaidVol*? |
|
Back to top |
|
|
pascuol n00b
Joined: 03 Dec 2006 Posts: 39
|
Posted: Sat Jul 21, 2012 3:17 pm Post subject: |
|
|
frostschutz wrote: | Hmmm, interesting. Not sure what the problem is.
mdadm -E /dev/RAID/RaidVol*? |
Code: | # mdadm -E /dev/RAID/RaidVol*
/dev/RAID/RaidVol1:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 9df9f556:ef8f3fb6:208753b0:486acf83
Name : cameleon:0 (local to host cameleon)
Creation Time : Fri Jan 13 11:48:23 2012
Raid Level : raid6
Raid Devices : 4
Avail Dev Size : 1133033472 (540.27 GiB 580.11 GB)
Array Size : 2266066944 (1080.54 GiB 1160.23 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : cb5e81cf:534d2c02:1da19216:394cec54
Update Time : Sat Jul 21 15:58:02 2012
Checksum : 21000f2b - correct
Events : 30330
Layout : left-symmetric
Chunk Size : 512K
Device Role : Active device 0
Array State : AAAA ('A' == active, '.' == missing)
/dev/RAID/RaidVol2:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 9df9f556:ef8f3fb6:208753b0:486acf83
Name : cameleon:0 (local to host cameleon)
Creation Time : Fri Jan 13 11:48:23 2012
Raid Level : raid6
Raid Devices : 4
Avail Dev Size : 1133033472 (540.27 GiB 580.11 GB)
Array Size : 2266066944 (1080.54 GiB 1160.23 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 25a8b562:eee9cfac:c865a211:46ed91f2
Update Time : Sat Jul 21 15:58:02 2012
Checksum : f78d5ad9 - correct
Events : 30330
Layout : left-symmetric
Chunk Size : 512K
Device Role : Active device 1
Array State : AAAA ('A' == active, '.' == missing)
/dev/RAID/RaidVol3:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 9df9f556:ef8f3fb6:208753b0:486acf83
Name : cameleon:0 (local to host cameleon)
Creation Time : Fri Jan 13 11:48:23 2012
Raid Level : raid6
Raid Devices : 4
Avail Dev Size : 1133033472 (540.27 GiB 580.11 GB)
Array Size : 2266066944 (1080.54 GiB 1160.23 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 45e96e83:26305401:f8d1bcbe:59b93147
Update Time : Sat Jul 21 15:58:02 2012
Checksum : 6e851a75 - correct
Events : 30330
Layout : left-symmetric
Chunk Size : 512K
Device Role : Active device 2
Array State : AAAA ('A' == active, '.' == missing)
/dev/RAID/RaidVol4:
Magic : a92b4efc
Version : 1.2
Feature Map : 0x0
Array UUID : 9df9f556:ef8f3fb6:208753b0:486acf83
Name : cameleon:0 (local to host cameleon)
Creation Time : Fri Jan 13 11:48:23 2012
Raid Level : raid6
Raid Devices : 4
Avail Dev Size : 2109790208 (1006.03 GiB 1080.21 GB)
Array Size : 2266066944 (1080.54 GiB 1160.23 GB)
Used Dev Size : 1133033472 (540.27 GiB 580.11 GB)
Data Offset : 2048 sectors
Super Offset : 8 sectors
State : clean
Device UUID : 33035e2f:5a0d2643:7acffbc8:1e03456c
Update Time : Sat Jul 21 15:58:02 2012
Checksum : c5d078df - correct
Events : 30330
Layout : left-symmetric
Chunk Size : 512K
Device Role : Active device 3
Array State : AAAA ('A' == active, '.' == missing)
|
|
|
Back to top |
|
|
frostschutz Advocate
Joined: 22 Feb 2005 Posts: 2977 Location: Germany
|
Posted: Sat Jul 21, 2012 5:34 pm Post subject: |
|
|
Hm, that's odd.
Did you do anything special for RaidVol4, cause that is the only one where it shows Code: | Avail Dev Size : 2109790208 (1006.03 GiB 1080.21 GB) |
Whereas for the others it says Code: | Avail Dev Size : 1133033472 (540.27 GiB 580.11 GB) |
So for some reason it's not detecting the new capacity of RaidVol1-3 but it's done so for 4.
Maybe it works after stopping & restarting the RAID? If all else fails you might have to remove, readd, resync each drive individually -- that's what you usually do anyway when replacing physical discs. |
|
Back to top |
|
|
pascuol n00b
Joined: 03 Dec 2006 Posts: 39
|
Posted: Sat Jul 21, 2012 8:38 pm Post subject: |
|
|
Doesn't change anything to unmount/remount or restarting the raid.
for the RaidVol4 I rebuilt it, I mean, remove, readd and resync.
I probably have to do it for each, but in my logic, I was thinking that I don't have to do it, and as it take 2 days per disks ... I prefer to avoid it.
is there any other way to tell the Raid that the volume size has changed ?
I will think about it tonight, and maybe try to remove / readd but without resyncing like readd with the option assumesafe (or something like that).
thanks for your help anyway, let me know if you have any good idea |
|
Back to top |
|
|
frostschutz Advocate
Joined: 22 Feb 2005 Posts: 2977 Location: Germany
|
Posted: Sat Jul 21, 2012 11:20 pm Post subject: |
|
|
You could try stopping the RAID, and then assembling it with the --update=devicesize option. This might force it to re-evaluate the size of its member drives.
Another option (but more dangerous) would be to create the RAID with the --assume-clean option. You'd have to use the same settings as your existing RAID upon create. Get the settings wrong and your data is corrupt. |
|
Back to top |
|
|
pascuol n00b
Joined: 03 Dec 2006 Posts: 39
|
Posted: Sun Jul 22, 2012 11:21 am Post subject: |
|
|
ok cool, it was the command I needed ..
Code: | mdadm -A /dev/md0 /dev/RAID/RaidVol[1-4] --update=devicesize
Size was 1133033472
Size is 2109790208
Size was 1133033472
Size is 2109790208
Size was 1133033472
Size is 2109790208
Size was 2109790208
Size is 2109790208
mdadm: /dev/md0 has been started with 4 drives.
|
and now it's resyncing
Code: | cat /proc/mdstat Sun Jul 22 13:01:17 2012
Personalities : [raid6] [raid5] [raid4]
md0 : active raid6 dm-0[4] dm-3[7] dm-2[6] dm-1[5]
2109790208 blocks super 1.2 level 6, 512k chunk, algorithm 2 [4/4] [UUUU]
[==========>..........] resync = 53.7% (566777192/1054895104) finish=1961.0min speed=4148K/sec
unused devices: <none>
|
it will "only" take 2000 min instead of much more but I'm already able to do the resize of my file system
So, now I've two method to resize my RAID, one is really slow, but keep the RAID up, with the other one, I need a short time to put it down .. Could be nice to be able to use the update=devicesize without stopping the RAID. |
|
Back to top |
|
|
|