raid problem

[brain salad surgery]

i used that link
http://www.finnie.org/text/terabyte/
and followed the instructions
(i have the same hardware)
all configs are fine, kernel, mdadm.conf and raidab

uname -r
2.6.17-gentoo-r8

hpt374 driver loads fine
Nov 6 13:57:25 mp3 scsi0 : hpt374
Nov 6 13:57:25 mp3 Vendor: HPT3xx Model: RAID 5 Array Rev: 3.00
Nov 6 13:57:25 mp3 Type: Direct-Access ANSI SCSI revision: 00
Nov 6 13:57:25 mp3 SCSI device sda: 960518784 512-byte hdwr sectors (491786 MB)
Nov 6 13:57:25 mp3 sda: Write Protect is off
Nov 6 13:57:25 mp3 sda: Mode Sense: 30 00 00 00
Nov 6 13:57:25 mp3 SCSI device sda: drive cache: write back
Nov 6 13:57:25 mp3 SCSI device sda: 960518784 512-byte hdwr sectors (491786 MB)
Nov 6 13:57:25 mp3 sda: Write Protect is off
Nov 6 13:57:25 mp3 sda: Mode Sense: 30 00 00 00
Nov 6 13:57:25 mp3 SCSI device sda: drive cache: write back
Nov 6 13:57:25 mp3 sda: sda1
Nov 6 13:57:25 mp3 sd 0:0:0:0: Attached scsi disk sda
Nov 6 13:57:25 mp3 Vendor: HPT3xx Model: RAID 5 Array Rev: 3.00
Nov 6 13:57:25 mp3 Type: Direct-Access ANSI SCSI revision: 00
Nov 6 13:57:25 mp3 SCSI device sdb: 960518784 512-byte hdwr sectors (491786 MB)
Nov 6 13:57:25 mp3 sdb: Write Protect is off
Nov 6 13:57:25 mp3 sdb: Mode Sense: 30 00 00 00
Nov 6 13:57:25 mp3 SCSI device sdb: drive cache: write back
Nov 6 13:57:25 mp3 SCSI device sdb: 960518784 512-byte hdwr sectors (491786 MB)
Nov 6 13:57:25 mp3 sdb: Write Protect is off
Nov 6 13:57:25 mp3 sdb: Mode Sense: 30 00 00 00
Nov 6 13:57:25 mp3 SCSI device sdb: drive cache: write back
Nov 6 13:57:25 mp3 sdb: sdb1
Nov 6 13:57:25 mp3 sd 0:0:1:0: Attached scsi disk sdb


more /etc/raidtab
raiddev /dev/md0
raid-level 0
nr-raid-disks 2
persistent-superblock 1
chunk-size 64
device /dev/sda1 # a raid5 array
raid-disk 0
device /dev/sdb1 # a raid5 array
raid-disk 1

All the partitions exist

more /proc/partitions
major minor #blocks name

3 0 6297480 hda
3 1 56196 hda1
3 2 506047 hda2
3 3 5735205 hda3
9 0 960510080 md0
8 0 480259392 sda
8 1 480255111 sda1
8 16 480259392 sdb
8 17 480255111 sdb1

mkraid /dev/md0
Nov 6 13:00:20 mp3 md: bind<sda1>
Nov 6 13:00:20 mp3 md: bind<sdb1>
Nov 6 13:00:20 mp3 md0: setting max_sectors to 128, segment boundary to 32767
Nov 6 13:00:20 mp3 raid0: looking at sdb1
Nov 6 13:00:20 mp3 raid0: comparing sdb1(480255040) with sdb1(480255040)
Nov 6 13:00:20 mp3 raid0: END
Nov 6 13:00:20 mp3 raid0: ==> UNIQUE
Nov 6 13:00:20 mp3 raid0: 1 zones
Nov 6 13:00:20 mp3 raid0: looking at sda1
Nov 6 13:00:20 mp3 raid0: comparing sda1(480255040) with sdb1(480255040)
Nov 6 13:00:20 mp3 raid0: EQUAL
Nov 6 13:00:20 mp3 raid0: FINAL 1 zones
Nov 6 13:00:20 mp3 raid0: done.
Nov 6 13:00:20 mp3 raid0 : md_size is 960510080 blocks.
Nov 6 13:00:20 mp3 raid0 : conf->hash_spacing is 960510080 blocks.
Nov 6 13:00:20 mp3 raid0 : nb_zone is 1.
Nov 6 13:00:20 mp3 raid0 : Allocating 4 bytes for hash.


mdadm -D /dev/md0
/dev/md0:
Version : 00.90.03
Creation Time : Mon Nov 6 13:00:20 2006
Raid Level : raid0
Array Size : 960510080 (916.01 GiB 983.56 GB)
Raid Devices : 2
Total Devices : 2
Preferred Minor : 0
Persistence : Superblock is persistent

Update Time : Mon Nov 6 13:00:20 2006
State : clean
Active Devices : 2
Working Devices : 2
Failed Devices : 0
Spare Devices : 0

Chunk Size : 64K

UUID : 495fd5c0:f1e4cee7:2ff5d4a4:dbd2f3aa
Events : 0.1

Number Major Minor RaidDevice State
0 8 1 0 active sync /dev/sda1
1 8 17 1 active sync /dev/sdb1


Now is the formatting that takes 15min but goes to the end:
mke2fs -j -m 0 -b 4096 -R stride=16 -O dir_index /dev/md0
Filesystem label=
OS type: Linux
Block size=4096 (log=2)
Fragment size=4096 (log=2)
120078336 inodes, 240127520 blocks
0 blocks (0.00%) reserved for the super user
First data block=0
Maximum filesystem blocks=243269632
7329 block groups
32768 blocks per group, 32768 fragments per group
16384 inodes per group
Superblock backups stored on blocks:
32768, 98304, 163840, 229376, 294912, 819200, 884736, 1605632, 2654208,
4096000, 7962624, 11239424, 20480000, 23887872, 71663616, 78675968,
102400000, 214990848
Writing inode tables: done
Creating journal (32768 blocks): done
Writing superblocks and filesystem accounting information: done

This filesystem will be automatically checked every 36 mounts or
180 days, whichever comes first. Use tune2fs -c or -i to override.



Then the problem:
mount /dev/md0 /home/mp3
mount: wrong fs type, bad option, bad superblock on /dev/md0,
missing codepage or other error
In some cases useful info is found in syslog - try
dmesg | tail or so

dmesg gives:
Nov 6 13:30:52 mp3 EXT3-fs error (device md0): ext3_check_descriptors: Block bitmap for group 2960 not in group (block 0)!
Nov 6 13:30:52 mp3 EXT3-fs: group descriptors corrupted !



what is it ? should i test for badblocks ?
i'm now running e2fsck -fy /dev/md0
which is finding a lot of blocks to reallocate,
inodes to recreate
how come on a newly formatted system ?

in fact the system is not new and i've done
all these steps a few times before without
that issue ...

[brain salad surgery]

now that e2fsck has been doing 3 times the scan pass1,
redoing the same things 3 times, i stopped it.

i did
raidstop -a
then
raidstart -a

more /proc/mdstat
Personalities : [linear] [raid0] [raid1] [raid10] [raid5] [raid4] [raid6] [multipath] [faulty]
unused devices: <none>

the logs:

Nov 6 15:11:32 mp3 raid0: looking at sda1
Nov 6 15:11:32 mp3 raid0: comparing sda1(480255040) with sda1(480255040)
Nov 6 15:11:32 mp3 raid0: END
Nov 6 15:11:32 mp3 raid0: ==> UNIQUE
Nov 6 15:11:32 mp3 raid0: 1 zones
Nov 6 15:11:32 mp3 raid0: FINAL 1 zones
Nov 6 15:11:32 mp3 raid0: too few disks (1 of 2) - aborting!
Nov 6 15:11:32 mp3 md: pers->run() failed ...
Nov 6 15:11:32 mp3 md: do_md_run() returned -12
Nov 6 15:11:32 mp3 md: md0 stopped.
Nov 6 15:11:32 mp3 md: unbind<sda1>
Nov 6 15:11:32 mp3 md: export_rdev(sda1)
Nov 6 15:11:32 mp3 md: ... autorun DONE.


well, looks like no change can be done on sdb1