| View previous topic :: View next topic |
| Author |
Message |
plink212
Tux's lil' helper
Joined: 21 Jan 2005
Posts: 90
|
 Posted: Tue Aug 24, 2010 6:09 pm Post subject: SOLVED: adding new drive to software raid array |
 |
|
I have 3 software raids setup
| Code: | Personalities : [raid0] [raid1] [raid10] [raid6] [raid5] [raid4]
md3 : active raid10 sdf2[1] sdb2[5] sda2[0] sde2[2] sdd2[3]
292675584 blocks 64K chunks 2 near-copies [6/5] [UUUU_U]
md4 : active raid5 sdf3[2] sdb3[5] sda3[0] sde3[3] sdd3[1]
6837411200 blocks level 5, 64k chunk, algorithm 2 [6/5] [UUUU_U]
md0 : active raid1 sdf1[2] sdb1[0] sda1[5] sde1[3] sdd1[4]
97536 blocks [6/5] [U_UUUU]
unused devices: <none>
|
as you can see one of my drives has failed and I need to replace it.
So I have a new drive which is /dev/sdc and I need to recreate the partition layout
This is the setup /dev/sda
| Code: | Model: ATA ST31500341AS (scsi)
Disk /dev/sda: 1500GB
Sector size (logical/physical): 512B/512B
Partition Table: gpt
Number Start End Size File system Name Flags
1 17.4kB 100MB 100MB ext2 primary raid
2 100MB 100GB 99.9GB xfs primary raid
3 100GB 1500GB 1400GB primary raid
|
I launched parted on /dev/sdc and issued
| Code: | (parted) mkpart primary 0 100M
Warning: The resulting partition is not properly aligned for best performance.
Ignore/Cancel? ignore
(parted) mkpart primary 100M 100G
(parted) mkpart primary 100G -1 |
but this ends up with
| Code: | Number Start End Size File system Name Flags
1 17.4kB 100MB 100MB ext2 primary
2 101MB 100GB 99.9GB primary
3 100GB 1500GB 1400GB primary
|
However when I try to add /dev/sdc2 to the array I get
| Code: | mdadm --manage /dev/md3 --add /dev/sdc2
mdadm: /dev/sdc2 not large enough to join array
|
I am confused why 2 starts at 101MB and not 100MB like the first.
Help
Last edited by plink212 on Tue Aug 24, 2010 6:30 pm; edited 1 time in total |
|
| Back to top |
|
 |
plink212
Tux's lil' helper
Joined: 21 Jan 2005
Posts: 90
|
| Posted: Tue Aug 24, 2010 6:30 pm Post subject: |
|
|
hmm if anyone else comes across this there is a workaround
in parted type
you can then specify the sector locations for the start and the end |
|
| Back to top |
|
|
Mad Merlin
Veteran
Joined: 09 May 2005
Posts: 1155
|
| Posted: Fri Aug 27, 2010 2:57 am Post subject: |
|
|
sfdisk makes it a lot easier to do this, as it can dump and recreate a partition table for you. Something like so:
| Code: | | sfdisk -d /dev/sda | sfdisk /dev/sdb |
Where /dev/sda is a drive currently working and part of the array, and /dev/sdb is a brand new, unpartitioned drive. Afterwards you can just add the newly created partitions to the arrays.
A more general problem that's easy to run into here is if you get another drive that's the same size, but has different geometry (usually happens with different brands or models), this often means you can't create partitions that are of identical size, and all of the new partitions must be at least as large as the existing ones. To get around this, try making your swap partition(s) smaller, even if you also have those in RAID arrays, it shouldn't be a problem to swapoff and recreate them.
_________________
Game! - Where the stick is mightier than the sword! |
|
| Back to top |
|
|
plink212
Tux's lil' helper
Joined: 21 Jan 2005
Posts: 90
|
| Posted: Fri Aug 27, 2010 5:25 am Post subject: |
|
|
agreed it would if they were mbr disks.
these were gpt disks and I couldn't get it to run against them
Tim |
|
| Back to top |
|
|
|
Kernel & Hardware
