Re: [Xen-users] Re: xen 2.0/3.0 pb with SW-RAID

[Fred Blaise <chapeaurouge@xxxxxxxxx>]

On 10/14/05, Marcus Brown <marcusbrutus@xxxxxxxxxxxxxxxx> wrote:
> Hi Fred,
>
> I'm not familiar with your hardware or distro, so hopefully I won't
> be adding to your problems!
>
> To me, these stand out:
>
>
> md: hdXn has different UUID to hdYm
>
> I'd have a look at making sure you have a valid mdadm.conf,
> then create a new initrd and see if things improve.
> I assume that the major difference between the two kernels (apart
> from one being xen-patched!) is that with the xen-patched one
> you've had to create your own initrd.
> AFAIK some mkinitrd methods need mdadm.conf so that mdadm
> (and others) are loaded properly.
> eg:
> # cat /etc/mdadm/mdadm.conf
> DEVICE partitions
> ARRAY /dev/md0 level=raid0 num-devices=2 
> UUID=72c1f448:2d5f8521:431d286d:3de5be1d
>    devices=/dev/hda3,/dev/hdc3
> ARRAY /dev/md1 level=raid0 num-devices=2 
> UUID=f739cdb1:7cfb3f92:605674de:969a9c53
>         devices=/dev/hda6,/dev/hdc6
> ARRAY /dev/md2 level=raid0 num-devices=2 
> UUID=0b81ec82:8cd3167c:8192ec98:f749babc
>         devices=/dev/hda7,/dev/hdc7
> # mdadm --examine /dev/hda3 | grep UUID
>            UUID : 72c1f448:2d5f8521:431d286d:3de5be1d
> # mdadm --examine /dev/hdc3 | grep UUID
>            UUID : 72c1f448:2d5f8521:431d286d:3de5be1d
I am using /etc/raidtab and the raidtools2, but I had noticed these
warnings as well. Will investigate.
>
>
> Also, to simplify things a little, you don't need to put your
> swap on a raid device, just give the two swap partitions the
> same priority. This will actually be faster too.
> eg:
> # grep swap /etc/fstab
> /dev/hda2               none                    swap    sw,pri=1        0     
>   0
> /dev/hdc2               none                    swap    sw,pri=1        0     
>   0
But if one of your disk crash, the swap on the good remaining disk
will not have the same info as the failed disk, right?
>
>
> The other msg that seems out of place is this:
> { DriveReady SeekComplete Error }
> and unless something really is wrong perhaps check DMA settings in
> your kernel config.
> grep DMA /mnt/kernels/config-2.6.11.12-xen0
> CONFIG_GENERIC_ISA_DMA=y
> CONFIG_BLK_DEV_IDEDMA_PCI=y
> # CONFIG_BLK_DEV_IDEDMA_FORCED is not set
> CONFIG_IDEDMA_PCI_AUTO=y
> # CONFIG_IDEDMA_ONLYDISK is not set
> CONFIG_BLK_DEV_IDEDMA=y
> # CONFIG_IDEDMA_IVB is not set
> CONFIG_IDEDMA_AUTO=y
>
To add more information on the hdd configuration I have, here is what
'raiddetect' says:
bicolor:/home/fblaise # raiddetect
The following disks hang off a HPT IDE controller:
The following disks hang off a PDC IDE controller:
name: /sys/block/hde, /dev name: /dev/hde, major: 33, minor: 0, size:
234441648, attached to: 0000:06:05.0, controller: 2, vendorid: 0x105a,
deviceid: 0x4d68, subsystem_vendorid: 0x105a, subsystem_deviceid:
0x4d68, controllertype: 2, hasraidmagic: 0
name: /sys/block/hdf, /dev name: /dev/hdf, major: 33, minor: 64, size:
234441648, attached to: 0000:06:05.0, controller: 2, vendorid: 0x105a,
deviceid: 0x4d68, subsystem_vendorid: 0x105a, subsystem_deviceid:
0x4d68, controllertype: 2, hasraidmagic: 0
name: /sys/block/hdg, /dev name: /dev/hdg, major: 34, minor: 0, size:
234441648, attached to: 0000:06:05.0, controller: 2, vendorid: 0x105a,
deviceid: 0x4d68, subsystem_vendorid: 0x105a, subsystem_deviceid:
0x4d68, controllertype: 2, hasraidmagic: 0
name: /sys/block/hdh, /dev name: /dev/hdh, major: 34, minor: 64, size:
234441648, attached to: 0000:06:05.0, controller: 2, vendorid: 0x105a,
deviceid: 0x4d68, subsystem_vendorid: 0x105a, subsystem_deviceid:
0x4d68, controllertype: 2, hasraidmagic: 0
The following disks hang off a SIL IDE controller:
The following disks are unclassified:
name: /sys/block/hda, /dev name: /dev/hda, major: 3, minor: 0, size:
234441648, attached to: 0000:00:1f.1, controller: 1, vendorid: 0x8086,
deviceid: 0x248b, subsystem_vendorid: 0x8086, subsystem_deviceid:
0x2480, controllertype: 0, hasraidmagic: 0
name: /sys/block/hdc, /dev name: /dev/hdc, major: 22, minor: 0, size:
234441648, attached to: 0000:00:1f.1, controller: 1, vendorid: 0x8086,
deviceid: 0x248b, subsystem_vendorid: 0x8086, subsystem_deviceid:
0x2480, controllertype: 0, hasraidmagic: 0
name: /sys/block/hdd, /dev name: /dev/hdd, major: 22, minor: 64, size:
8388604, attached to: 0000:00:1f.1, controller: 1, vendorid: 0x8086,
deviceid: 0x248b, subsystem_vendorid: 0x8086, subsystem_deviceid:
0x2480, controllertype: 0, hasraidmagic: 0

Thanks.
>
> Marcus.
fred
>

_______________________________________________
Xen-users mailing list
Xen-users@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-users