After emerging baselayout-1.11.3 it is no longer possible to boot into the system when using /dev/mdx as the root partition. The ro phase of the boot process seems to go on correctly but the failure occurs during an attempt to check the filesystem. I use reiserfs and reiserfsck complains that it does not see the root device. The boot process stops at this point with a prompt saying to either press CTRL-D to continue (that is to reboot) or to type-in the root password. Having done the latter it is possible to browse the root file system - apparently it is mounted read only. I do not use mdadm but raidtools instead. My raidtab looks as the following: raid-level 1 nr-raid-disks 2 chunk-size 32 persistent-superblock 1 device /dev/hde1 raid-disk 0 device /dev/hdg1 raid-disk 1 raiddev /dev/md1 raid-level 1 nr-raid-disks 2 chunk-size 32 persistent-superblock 1 device /dev/hde2 raid-disk 0 device /dev/hdg2 raid-disk 1 raiddev /dev/md2 raid-level 1 nr-raid-disks 2 chunk-size 32 persistent-superblock 1 device /dev/hde3 raid-disk 0 device /dev/hdg3 raid-disk 1 raiddev /dev/md3 raid-level 1 nr-raid-disks 2 chunk-size 32 persistent-superblock 1 device /dev/hda1 raid-disk 0 device /dev/hdc1 raid-disk 1 doing: # ls /dev reveals that only md0 is present as a softlink : /dev/md0 -> /dev/md/0 no /dev/md1 or /dev/md2 can be found. Downgrading to sys-apps/baselayout-1.10.4 and performing etc-update in order to undo the changes solves the problem. Of course it is not as easy as that. You must get into the defunct system first so let me describe what I did - maybe this will be helpful for someone who has encountered the same problem. Boot from a live CD and then: # modprobe raid1 # nano -w /etc/raidtab type-in the raidtable as listed above # raidstart /dev/md2 # mount /dev/md2 /mnt/gentoo # mount -t proc none /mnt/gentoo/proc # chroot /mnt/gentoo /bin/bash # env-update && source /etc/profile # echo "=sys-apps/baselayout-1.11.3" >> /etc/portage/package.mask (I chose the above solution to get rid of the unwanted version of the package for good - this ensures it will not be reemerged in the future) # unlink /dev/eth0 the above step is necessary because in the new layout eth0 is a softlink pointing at /dev/lo, in previous versions it was an independent file. If you fail to do that your network devices will not work after reemerging an older layout. # emerge baselayout at this point sys-apps/baselayout-1.10.4 got emerged # etc-update perform configuration files update. # exit # unmount /mnt/gentoo/proc /mnt/gentoo # reboot (do not forget to remove the CD) Possibly this is just a misunderstandig and the new baselayout works well with mdadm - I have not checked that. If raidtools and raidtab are deprecated please let me know how to switch to mdamd safely that is without damaging my existing raid partitions and rebuilding the whole system from scratch. Reproducible: Always Steps to Reproduce: 1. # emerge "=sys-apps/baselayout-1.11.3" 2. # etc-update 3. update the configuration files to the best of my will and knowledge :-) 4. # reboot Actual Results: The system failed to finish booting with a claim that reiserfsck could not see the root device (in my case /dev/md2). Expected Results: The system should have booted-up as always.
are you using devfs ? if so, please try 1.11.4
Yes soft raid works correctly again with baselayout-1.11.4. Thank you for an extremely fast reaction to this bug. I wish Microsoft(TM) behaved that way:-) Answering the previous post - yes I am still using devfs. I have not switched to udev yet.
just glad it boots for you ;) *** This bug has been marked as a duplicate of 68795 ***
