| View previous topic :: View next topic |
| Author |
Message |
eddy89
Apprentice
Joined: 01 Feb 2006
Posts: 180
Location: /world/Italy/Torino
|
 Posted: Thu May 21, 2009 4:32 pm Post subject: raid=part ignored after updating to 2.6.28-gentoo-r5 |
 |
|
In my headless server configuration I have raid1 configuration:
| Code: |
cat /proc/mdstat
Personalities : [raid1]
md_d3 : active raid1 hdc2[1] hda1[0]
80035712 blocks [2/2] [UU]
|
This device has partition, and with this kernel command line:
| Code: | | kernel /boot/kern root=/dev/md_d3p1 ide0=ata66 ide1=ata66 raid=part |
With kernel 2.6.24-gentoo-r8 i get to autodetect and boot mounting one partition in the raid device:
| Code: | md: Autodetecting RAID arrays.
md: Scanned 2 and added 2 devices.
md: autorun ...
md: considering hdc2 ...
md: adding hdc2 ...
md: adding hda1 ...
md: created md_d3
md: bind<hda1>
md: bind<hdc2>
md: running: <hdc2><hda1>
raid1: raid set md_d3 active with 2 out of 2 mirrors
md: ... autorun DONE.
md_d3: p1 p2 p3 < p5 p6 p7 >
kjournald starting. Commit interval 5 seconds
EXT3-fs: mounted filesystem with ordered data mode.
VFS: Mounted root (ext3 filesystem) readonly.
Freeing unused kernel memory: 184k freed
EXT3 FS on md_d3p1, internal journal
kjournald starting. Commit interval 5 seconds
|
Then I updated to 2.6.28-gentoo-r5 and at boot it doesn't recognise partition, something like this:
| Code: | md: Autodetecting RAID arrays.
md: Scanned 2 and added 2 devices.
md: autorun ...
md: considering hdc2 ...
md: adding hdc2 ...
md: adding hda1 ...
md: created md_d3
md: bind<hda1>
md: bind<hdc2>
md: running: <hdc2><hda1>
raid1: raid set md_d3 active with 2 out of 2 mirrors
md: ... autorun DONE. |
Then panic because it can't mount the root partition.
Now did I mistaken something or it is a bug I should report?
(The new config is made using make oldconfig).
Do you have any solution?
I couldn't find anything about this on the net, could you help me?. |
|
| Back to top |
|
 |
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54507
Location: 56N 3W
|
| Posted: Thu May 21, 2009 7:12 pm Post subject: |
|
|
eddy89,
Your raid is detected, | Code: | md: running: <hdc2><hda1>
raid1: raid set md_d3 active with 2 out of 2 mirrors
md: ... autorun DONE. |
but the ext3fs it contains is not.
This suggests that ext3 is not available in the kernel. It must be <*> not
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
eddy89
Apprentice
Joined: 01 Feb 2006
Posts: 180
Location: /world/Italy/Torino
|
| Posted: Thu May 21, 2009 8:02 pm Post subject: |
|
|
The raid is detected but not the partitions in the raid:
| Code: | md: ... autorun DONE.
md_d3: p1 p2 p3 < p5 p6 p7 > |
The ext3 is on md_d3p1, if the kernel doesn't have this device it couldn't mount it ^^.
The problem is that the partition table is not scanned also if there is the option "raid=part".
from the man page md(4):
| Code: | raid=partitionable
raid=part
These are available in 2.6 and later kernels only. They indicate
that autodetected MD arrays should be created as partitionable
arrays, with a different major device number to the original non-
partitionable md arrays. The device number is listed as mdp in
/proc/devices.
|
|
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54507
Location: 56N 3W
|
| Posted: Thu May 21, 2009 8:05 pm Post subject: |
|
|
eddy89,
I missed that. I use lvm on raid, so I've not tried partitioning kernel raid devices.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
eddy89
Apprentice
Joined: 01 Feb 2006
Posts: 180
Location: /world/Italy/Torino
|
| Posted: Thu May 21, 2009 9:20 pm Post subject: |
|
|
I tryed to reproduce the problem in my (non production) desktop, and it is reproducible:
If I force the auto detection with "mdadm --auto-detect" it works:
| Code: | md: Autodetecting RAID arrays.
md: Scanned 2 and added 2 devices.
md: autorun ...
md: considering sdb2 ...
md: adding sdb2 ...
md: adding sdb1 ...
md: created md0
md: bind<sdb1>
md: bind<sdb2>
md: running: <sdb2><sdb1>
md0: WARNING: sdb1 appears to be on the same physical disk as sdb2.
True protection against single-disk failure might be compromised.
raid1: raid set md0 active with 2 out of 2 mirrors
md: ... autorun DONE.
md0: p1 p3 |
(Note the last line)
If I try to boot with root=/dev/md0p1, it starts and success the auto detection, but the last line doesn't appear and the kernel panic.
There is something I can do to resolve this problem, without downgrading the kernel?? Can I know if the problem will be solved in new versions?
Should I post a bug? |
|
| Back to top |
|
|
NeddySeagoon
Administrator
Joined: 05 Jul 2003
Posts: 54507
Location: 56N 3W
|
| Posted: Thu May 21, 2009 9:44 pm Post subject: |
|
|
eddy89,
Check bugs.gentoo.org and post a bug there if its new.
Add to an existing bug if you can provide more information than is already on the bug.
The Gentoo kernel devs will flow it upstream if required.
_________________
Regards,
NeddySeagoon
Computer users fall into two groups:-
those that do backups
those that have never had a hard drive fail. |
|
| Back to top |
|
|
eddy89
Apprentice
Joined: 01 Feb 2006
Posts: 180
Location: /world/Italy/Torino
|
|
| Back to top |
|
|
|
Kernel & Hardware
