Red Hat Bugzilla – Bug 91738
upgrade with raid/lvm fails with wrong root and/or fstab error
Last modified: 2007-04-18 12:54:03 EDT
Description of problem:The installation tries to mount the partions but seems to use a wrong partion as '/', the right '/' would be '/dev/vg_system/root', it uses '/dev/md0' first which is '/boot'.It complains about errors with fstab and stops the upgrade process.I tried to upgrade from redhat-8.0. The LVM is on /dev/md1.I can manually mount /dev/vg_system/root and after that chroot with 'mount -av' works. I tried a kickstart installation as upgrade but with no success, too./etc/fstab:/dev/vg_system/root / ext2 defaults 1 1/dev/md0 /boot ext2 defaults 1 2none /dev/pts devpts gid=5,mode=620 0 0/dev/vg_system/home /home ext3 defaults 1 2/dev/vg_system/opt /opt ext3 defaults 1 2none /proc proc defaults 0 0none /dev/shm tmpfs defaults 0 0/dev/vg_system/tmp /tmp ext2 defaults 1 2/dev/vg_system/usr /usr ext3 defaults 1 2/dev/vg_system/local /usr/local ext3 defaults 1 2/dev/vg_system/var /var ext3 defaults 1 2/dev/sdb2 swap swap defaults 0 0/dev/sda2 swap swap defaults 0 0/dev/fd0 /mnt/floppy auto noauto,owner,kudzu 0 0/dev/cdrom /mnt/cdrom iso9660 noauto,owner,kudzu,ro 0 0Version-Release number of selected component (if applicable):How reproducible:Steps to Reproduce:1.install redhat-8.0 setup like above2.try upgrade Actual results:during upgrade error about fstab, mount failes and upgrade stoppedExpected results:Additional info:anaconda.log:[...]* Running kickstart %pre script(s)* All kickstart %pre script(s) have been run* moving (1) to step findrootparts* going to mount /tmp/md0 on /mnt/sysimage* going to mount /dev/vg_system/var on /mnt/sysimage* going to mount /dev/vg_system/local on /mnt/sysimage* going to mount /dev/vg_system/usr on /mnt/sysimage* going to mount /dev/vg_system/tmp on /mnt/sysimage* going to mount /dev/vg_system/opt on /mnt/sysimage* going to mount /dev/vg_system/home on /mnt/sysimage* going to mount /dev/vg_system/root on /mnt/sysimage* moving (1) to step partitionobjinit* moving (1) to step upgrademount* going to mount /dev/vg_system/root on /mnt/sysimage as ext2* going to mount /dev/vg_system/root on /mnt/sysimage
Created attachment 92007 [details]
Created attachment 92008 [details]
When asked about which installation(s) you would like to upgrade which one(s)
are you presented with? From the log it looks like we are trying to mount
/dev/vg_system/root as the last thing we do.
It was 'Redhat Linux 8.0 (/dev/vg_system/root)', the only choise. I tried in
english upgrade now (german before). The requester have the title 'Mount failed'
and the text 'One or more of the filesystems listed in the /etc/fstab on your
linux system cannot be mountet. ...'.
When this error occurs look on VC3 (cntl-alt-f3) and see if you can find any log
output related to the failure to mount the partition.
At 'VC3' I see exactly the same as in the attached 'anaconda.log'.
At 'VC4' I messages from kernel and other components, which is the same as im
In both I don't see any error messages from 'mount'.
I attach the 'syslog' now, too.
Created attachment 92074 [details]
If you run 'e2label /dev/md0' what does it say the label is?
The label of /dev/md0 is '/boot'. There is no 'LABEL=' used in /etc/fstab, there
are only device files or 'none'.
Fix for this in CVS.