On Tue, Jan 26, 2010 at 07:06:26PM +0100, Ferenc Wagner wrote: > Package: kpartx > Version: 0.4.8-14+lenny2 > Severity: normal > > Hi, > > I'd like to boot from a partitioned multipath device, but the device nodes > corresponding to the partitions aren't created. At the (initramfs) prompt: > > (initramfs) multipath -ll > nfs (360060160ec612000f6f237e09406df11) dm-0 DGC ,RAID 5 > [size=5.0G][features=1 queue_if_no_path][hwhandler=1 emc] > \_ round-robin 0 [prio=2][active] > \_ 1:0:1:0 sdb 8:16 [active][ready] > \_ 0:0:1:0 sdd 8:48 [active][ready] > \_ round-robin 0 [prio=0][enabled] > \_ 1:0:0:0 sda 8:0 [active][ready] > \_ 0:0:0:0 sdc 8:32 [active][ready] > > which looks perfectly good, but > > (initramfs) ls -l /dev/disk/by-id > lrwxrwxrwx 1 0 0 16 Jan 26 17:38 dm-name-nfs -> > ../../mapper/nfs > lrwxrwxrwx 1 0 0 16 Jan 26 17:38 > dm-uuid-mpath-360060160ec612000f6f237e09406df11 -> ../../mapper/nfs > lrwxrwxrwx 1 0 0 9 Jan 26 17:50 > scsi-360060160ec612000f6f237e09406df11 -> ../../sda > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > scsi-360060160ec612000f6f237e09406df11-part1 -> ../../sdd1 > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > scsi-360060160ec612000f6f237e09406df11-part2 -> ../../sdd2 > lrwxrwxrwx 1 0 0 9 Jan 26 17:38 > scsi-SIBM-ESXSST936701LC_F3LC065GL00007604NMST -> ../../sde > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > scsi-SIBM-ESXSST936701LC_F3LC065GL00007604NMST-part1 -> ../../sde1 > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > scsi-SIBM-ESXSST936701LC_F3LC065GL00007604NMST-part2 -> ../../sde2 > lrwxrwxrwx 1 0 0 9 Jan 26 17:50 > wwn-0x60060160ec612000 -> ../../sda > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > wwn-0x60060160ec612000-part1 -> ../../sdd1 > lrwxrwxrwx 1 0 0 10 Jan 26 17:38 > wwn-0x60060160ec612000-part2 -> ../../sdd2 > > which feels rather confused (sd[a-d] are the paths, sde is a local disk). > Basically, I miss dm-name-nfs-part[12] and the corresponding > /dev/mapper/nfs-part[12] > devices. > > (initramfs) kpartx -l /dev/mapper/nfs > nfs1 : 0 385497 /dev/mapper/nfs 63 > nfs2 : 0 10088820 /dev/mapper/nfs 385560 > > which is about right, kpartx -a /dev/mapper/nfs; vgchange -ay nfs lets the > boot > conclude normally (/dev/mapper/nfs2 is my PV), although /dev/mapper/nfs1 > disappears, so later I get > > Checking file systems...fsck from util-linux-ng 2.16.2 > fsck.ext2: Bad magic number in super-block while trying to open > /dev/mapper/nfs1 > > but that's probably the same issue striking again. > /lib/udev/rules.d/60-kpartx.rules is in place and should work, but somehow it > does not. At least the above device/partition mixup is not present in the > booted > system: > > lrwxrwxrwx 1 root root 9 2010-01-26 18:58 > scsi-360060160ec612000f6f237e09406df11 -> ../../sdd > lrwxrwxrwx 1 root root 10 2010-01-26 18:57 > scsi-360060160ec612000f6f237e09406df11-part1 -> ../../sdd1 > lrwxrwxrwx 1 root root 10 2010-01-26 18:57 > scsi-360060160ec612000f6f237e09406df11-part2 -> ../../sdd2 > lrwxrwxrwx 1 root root 9 2010-01-26 18:58 wwn-0x60060160ec612000 -> ../../sdd > lrwxrwxrwx 1 root root 10 2010-01-26 18:57 wwn-0x60060160ec612000-part1 -> > ../../sdd1 > lrwxrwxrwx 1 root root 10 2010-01-26 18:57 wwn-0x60060160ec612000-part2 -> > ../../sdd2 > > Interestingly, > > lrwxrwxrwx 1 root root 17 2010-01-26 18:57 /dev/disk/by-id/dm-name-nfs1 -> > ../../mapper/nfs1 > > is also present, although broken in the booted system. > Udev changed a lot since Lenny, I'd appreciate some help finding the problem. Could you add a "sleep 10" at the end of the multipath local-top initramfs script? It's possible that lvm "grabs" the sd* devices before multipath had a chance to assemble the multipath maps. That would at least look like the above. Cheers, -- Guido
-- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
