isw: unsupported map state 0x3 on /dev/sda for RAID_Volume0
Christian Schuhegger
Christian.Schuhegger at gmx.de
Sun Jun 8 16:37:26 UTC 2008
Hello,
I am running ubuntu/windows vista in a dual boot set-up on a machine
with an intel matrix raid controller. I started to do that with gutsy
7.10 already last year in June and upgraded a month ago to hardy 8.04.
After the upgrade I had some difficulties, because ubuntu does not
support fake raids by default, but I got it under control.
https://help.ubuntu.com/community/FakeRaidHowto
Since yesterday the system refuses to boot. The initial bios screen
tells me that the raid volume 0 has failed and finally I only get the
message:
"disk boot failure, insert system disk and press enter"
Currently I try to analyse the problem with a KNOPPIX 5.3.1 and dmraid.
I've attached the output of
dmraid -ay -vvvv -dddd
and
dmraid -n
I've also generated the output of
dmraid -r -D /dev/sda
dmraid -r -D /dev/sdb
One of my questions is, if the metadata of both disks should be
identical? Becaus if I create a "hexdump -C" and run diff on the output
I get differences (attached). Another question would be if I modify the
metadata by hand and write it back to the disk if this would work,
because I see in the data structures defined in:
dmraid/1.0.0.rc14/lib/format/ataraid/isw.h
-- snip start --
struct isw {
int8_t sig[MAX_SIGNATURE_LENGTH];/* 0x0 - 0x1F */
uint32_t check_sum; /* 0x20 - 0x23 MPB Checksum */
uint32_t mpb_size; /* 0x24 - 0x27 Size of MPB */
uint32_t family_num; /* 0x28 - 0x2B Checksum from
first time this config was written */
-- snip end --
that there are some checksums.
When I run
fdisk -l /dev/sda
I get the following output:
unable to seek on /dev/sda
On /dev/sdb fdisk starts without problems.
After my experiments with knoppix and dmraid finally the system is
booting again, even if I did not write anything back on the disks, at
least not knowingly. Windows does not show any strange behaviour up to now.
I am uncertain if the problem is due to a real hardware problem or due
to some misbehaviour of Linux, especially, because I found hints that
ubuntu hardy 8.04 has some problems in that direction:
https://bugs.launchpad.net/ubuntu/+source/dmraid/+bug/219393
Another bug report led me to upgrade to kernel 2.6.24-18 just before my
problems with the raid started:
https://bugs.launchpad.net/ubuntu/+source/linux/+bug/209454
I would be interested in your opinion on how to continue from here. I've
ordered a big external harddrive to at least backup the full disk
content, but after that backup what would be the recommended way to
continue? Replace the disks, because it is a hardware problem or fix
ubuntu, because some bug in ubuntu may cause the problem?
Many thanks and have a nice day,
--
Christian Schuhegger
http://www.el-chef.de/
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: dmraid_ay_v.txt
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: dmraid_n.txt
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment-0001.txt>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: sda_isw.dat.hex
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment.ksh>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: sdb_isw.dat.hex
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment-0001.ksh>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: sd_isw.dat.hex.diff
Type: text/x-diff
Size: 3476 bytes
Desc: not available
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment.bin>
-------------- next part --------------
An embedded and charset-unspecified text was scrubbed...
Name: chipset.txt
URL: <http://listman.redhat.com/archives/ataraid-list/attachments/20080608/9c8fd949/attachment-0002.txt>
More information about the Ataraid-list
mailing list