PDA

View Full Version : FC6 + RAID5 woes



capo
30th October 2006, 07:09 AM
I just did a fresh isntall of fc6 after completely wiping all three of my seagate baracude 160bg drives. I used the partitioner in the installer to create the following raid arrays:

100mb RAID 1 /boot
300GB RAID 5 /
2.1gb RAID 0 linux swap

Now boot and swap work fine, but while booting i get the error msg:

md1 raid level 5 working with 2 of 3 devices active, using algorithem 1.

Along those lines anyway. Now all three drives are obviously working, i maniged to farmat then fine and my swap raid 0 is obviously working fine and if there was a single disk not up the raid 0 would fail completely and i would know about it. So why is my raid 5 not working. Does anyone know how i can fix this.

NOTE: i also had the same issue with fc5 on this hardware

MY HARDWARE IS:

Core duo 2 e6400
gigabyte p965 ds3
2gb G.Skill HZ
3 160GB Hdisks

Thanks in advance, capo.

capo
31st October 2006, 04:24 AM
No one else had issues with raid 5 and fedora core on gigabyte ds3?

capo
1st November 2006, 11:11 AM
ok well obviously no one know wtf is going on here so does anyone atleast know where i should be posting this to get some help?

huw-l
1st November 2006, 09:19 PM
I think you may only be able to use RAID1 for /

what do you get from 'cat /proc/mdstat'

capo
2nd November 2006, 01:21 AM
cat /proc/mdstat:

md0 : active raid1 hdc1[1] hda1[0]
104320 blocks [3/2] [UU_]

md2 : active raid0 sda3[2] hdc3[1] hda3[0]
2216448 blocks 256k chunks

md1 : active raid5 hdc2[1] hda2[0]
310889472 blocks level 5, 256k chunk, algorithm 2 [3/2] [UU_]

huw-l
2nd November 2006, 07:34 PM
and 'fdisk -l' ?

why is md2 on sda3 and not hda3? have you got SATA and paralell ATA drives in this machine?

capo
3rd November 2006, 12:42 AM
nope there all sata drivers in the raid. I do have a pata drive but it's not at all involved. I think it must just be confused with the names. Should i perhaps remove the pata driver and reformat and reconfigure the raid with it not in there?

capo
3rd November 2006, 04:00 PM
[root@certain-beast ~]# fdisk -l

Disk /dev/hdf: 40.0 GB, 40020664320 bytes
255 heads, 63 sectors/track, 4865 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/hdf1 * 1 4864 39070048+ 7 HPFS/NTFS

Disk /dev/hda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/hda1 * 1 13 104391 fd Linux raid autodetect
/dev/hda2 14 19365 155444940 fd Linux raid autodetect
/dev/hda3 19366 19457 738990 fd Linux raid autodetect

Disk /dev/hdc: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/hdc1 * 1 13 104391 fd Linux raid autodetect
/dev/hdc2 14 19365 155444940 fd Linux raid autodetect
/dev/hdc3 19366 19457 738990 fd Linux raid autodetect

Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 * 1 13 104391 fd Linux raid autodetect
/dev/sda2 14 19365 155444940 fd Linux raid autodetect
/dev/sda3 19366 19457 738990 fd Linux raid autodetect

Disk /dev/md1: 318.3 GB, 318350819328 bytes
2 heads, 4 sectors/track, 77722368 cylinders
Units = cylinders of 8 * 512 = 4096 bytes

Disk /dev/md1 doesn't contain a valid partition table

Disk /dev/md2: 2269 MB, 2269642752 bytes
2 heads, 4 sectors/track, 554112 cylinders
Units = cylinders of 8 * 512 = 4096 bytes

Disk /dev/md2 doesn't contain a valid partition table

Disk /dev/md0: 106 MB, 106823680 bytes
2 heads, 4 sectors/track, 26080 cylinders
Units = cylinders of 8 * 512 = 4096 bytes

Disk /dev/md0 doesn't contain a valid partition table

capo
4th November 2006, 02:22 AM
Ok so looking @ my fdisk -l and my cat /proc/mdstat is there anyway to add the missing HD back in and rebuild the arrays?

capo
5th November 2006, 07:53 AM
Ok so reformatting with the other HD out did not help, nor did reformattin in various other configs with differnt HD's in differnt ports etc. Alwayays the same disk drops out fo the raid 5 and the raid 1. but works fine for raid 0. I treid adding the disk back into the raid 5 adn the raid 1, in both cases it added but showed up as failed:

[capo@certain-beast ~]$ cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid0] [raid1]
md0 : active raid1 sdb1[3](F) hda1[0] sda1[1]
104320 blocks [3/2] [UU_]

md1 : active raid0 sdb2[2] sda2[1] hda2[0]
2433024 blocks 256k chunks

md2 : active raid5 sdb3[3](F) sda3[1] hda3[0]
310745088 blocks level 5, 256k chunk, algorithm 2 [3/2] [UU_]

Now clearly the disk itself is Mostly OK. I can format it with fdisk, i can create filesystems on it ok with mk2fs i can run a perfectly stable raid 0 using a partition on it. But raid 1 and 5 dont like it.

firstly yes it is IDENTICLE to the other 2 HD's. There all exactly the same model. So something must be wrong with the way it is formatted or some such. Can anyone think of what could be wrong our is there some tool i can use to completely scrub it, delete everything, actualy write a 0 to every bit including the mbr and write a new empty MBR onto it. Would this even help? I can think o fno other reason something like raid0 could work but not raid 1 or 5.

nick.stumpos
5th November 2006, 08:09 AM
dd if=/dev/zero of=/dev/hdX will write all zeros

capo
7th November 2006, 02:56 AM
well i did that then repartioned then tried to rebuild the arrays. Raid 0 still works. Raid 1 and 5 it still comes up as a failed disk. Even though i had no problems partitioning it or writing lots of zeroes 2 it.

huw-l
7th November 2006, 03:21 PM
so looks like it's getting confused by the raid autodetect partitions on the sata drive. I suspect if you remove the SATA drive or change it's partition types to something other than "raid autodetect" the arrays should sort themselves out on reboot.

If not you can use the mdadm command to add disks back into arrays.

capo
8th November 2006, 02:55 AM
i already tried that, see let me quote myself from my second last post.

"Ok so reformatting with the other HD out did not help, nor did reformattin in various other configs with differnt HD's in differnt ports etc. Alwayays the same disk drops out fo the raid 5 and the raid 1. but works fine for raid 0. I treid adding the disk back into the raid 5 adn the raid 1, in both cases it added but showed up as failed:

[capo@certain-beast ~]$ cat /proc/mdstat
Personalities : [raid6] [raid5] [raid4] [raid0] [raid1]
md0 : active raid1 sdb1[3](F) hda1[0] sda1[1]
104320 blocks [3/2] [UU_]

md1 : active raid0 sdb2[2] sda2[1] hda2[0]
2433024 blocks 256k chunks

md2 : active raid5 sdb3[3](F) sda3[1] hda3[0]
310745088 blocks level 5, 256k chunk, algorithm 2 [3/2] [UU_]

Now clearly the disk itself is Mostly OK. I can format it with fdisk, i can create filesystems on it ok with mk2fs i can run a perfectly stable raid 0 using a partition on it. But raid 1 and 5 dont like it.

firstly yes it is IDENTICLE to the other 2 HD's. There all exactly the same model. So something must be wrong with the way it is formatted or some such. Can anyone think of what could be wrong our is there some tool i can use to completely scrub it, delete everything, actualy write a 0 to every bit including the mbr and write a new empty MBR onto it. Would this even help? I can think o fno other reason something like raid0 could work but not raid 1 or 5."