Hi Jane. On Wed, Aug 18, 2004 at 02:07:49PM -0700, Jane Liu wrote: > Hi, > > This is in a way continuation of the discussion at > https://www.redhat.com/archives/ataraid-list/2004-August/msg00013.html. > > Just tried version 1.0.0-rc3 on my ICH5-R system, and still couldn't > discover the array. Well, I didn't get arround to your metadata yet ;) > The problem I had with 1.0.0-rc2 turned out to be > caused by bogus disk geometry data ( i guess [heads * sectors * > cylinders] does not necessarily equal the total sector count). Yes, 2.6 is still bogus in that regard. > I noticed > that in 1.0.0-rc3 the total sector count is not computed from geometry > data but read from sysfs. Which was the case in 1.0.0-rc2 as well presumably sysfs was accessibale. > So now I always end up with a 0 sector count > because I'm not using sysfs. As a result, I get errors seeking to the > superblock offset and couldn't find the metadata. > > So is dmraid intended for use with sysfs only now? If not, how can I get > it to obtain the correct sector count when "carrying on with /dev/"? I > mean, should I do something outside of dmraid to achieve this? Yes, sysfs is meant to be used in 2.6. Is there a special reason, why you can't use it ? If so, I need to think about the ioctl workaround you mention below for such configurations. Will try to get around to your metadata copies tomorrow... > > For now, I'm getting around this problem by seeking from the end of the > device file to get to the superblock (since I end up with negative > superblock offsets). This is kind of an ugly hack. Alternatively, it is > also possible to get the sector count using the BLKGETSIZE ioctl, which > is what hdparm uses. I don't know enough about this method to advocate > for or against it, but it works on my setup. > > Sorry if I'm being over-zealous :-) I need to return the ICH5-R to our > QA team soon, and I really want to get as much testing done as I can. > > Here's output from "dmraid -r -vvv" with a couple of added prints: > > INFO: locking /var/lock/dmraid/.lock > ERROR: Failed to find sysfs mount point > carrying on with /dev/ > INFO: /dev/sdb: hpt37x discovering > INFO: /dev/sdb: hpt45x discovering > ERROR: failed seeking device "/dev/sdb" at offset -5632 > INFO: /dev/sdb: isw discovering > isw_sboffset is: 0 > ERROR: failed seeking device "/dev/sdb" at offset -1024 > ERROR: isw: failed reading first sector of ISW superblock for /dev/sdb > INFO: /dev/sdb: lsi discovering > ERROR: failed seeking device "/dev/sdb" at offset -512 > INFO: /dev/sdb: pdc discovering > ERROR: failed seeking device "/dev/sdb" at offset -32256 > INFO: /dev/sdb: sil discovering > ERROR: failed seeking device "/dev/sdb" at offset -512 > INFO: /dev/sda: hpt37x discovering > INFO: /dev/sda: hpt45x discovering > ERROR: failed seeking device "/dev/sda" at offset -5632 > INFO: /dev/sda: isw discovering > isw_sboffset is: 0 > ERROR: failed seeking device "/dev/sda" at offset -1024 > ERROR: isw: failed reading first sector of ISW superblock for /dev/sda > INFO: /dev/sda: lsi discovering > ERROR: failed seeking device "/dev/sda" at offset -512 > INFO: /dev/sda: pdc discovering > ERROR: failed seeking device "/dev/sda" at offset -32256 > INFO: /dev/sda: sil discovering > ERROR: failed seeking device "/dev/sda" at offset -512 > INFO: /dev/hda: hpt37x discovering > INFO: /dev/hda: hpt45x discovering > ERROR: failed seeking device "/dev/hda" at offset -5632 > INFO: /dev/hda: isw discovering > isw_sboffset is: 0 > ERROR: failed seeking device "/dev/hda" at offset -1024 > ERROR: isw: failed reading first sector of ISW superblock for /dev/hda > INFO: /dev/hda: lsi discovering > ERROR: failed seeking device "/dev/hda" at offset -512 > INFO: /dev/hda: pdc discovering > ERROR: failed seeking device "/dev/hda" at offset -32256 > INFO: /dev/hda: sil discovering > ERROR: failed seeking device "/dev/hda" at offset -512 > No Software RAID disks > > Cheers, > Jane > _______________________________________________ > > Ataraid-list@xxxxxxxxxx > https://www.redhat.com/mailman/listinfo/ataraid-list -- Regards, Heinz -- The LVM Guy -- *** Software bugs are stupid. Nevertheless it needs not so stupid people to solve them *** =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- Heinz Mauelshagen Red Hat GmbH Consulting Development Engineer Am Sonnenhang 11 56242 Marienrachdorf Germany Mauelshagen@xxxxxxxxxx +49 2626 141200 FAX 924446 =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-
