Red Hat Bugzilla – Bug 220096
Disk partitioning is set to LVM2 on the default installation, however LVM2 is not supported on the rescue CD
Last modified: 2013-02-28 23:04:24 EST
Description of problem: By default, when you perform a basic installation of Core 6 the basic partitioning schema that diskdruid uses is an LVM. The problem is in that the Rescue CD does not have LVM support built into the kernel and none of the associated LVM tools (which could be used for recovery and the like) are available. This makes the Rescue CD functionally useless for 100% of all default installations. Version-Release number of selected component (if applicable): How reproducible: Very reproducable, the problem is with the way the rescue CD is packaged. Steps to Reproduce: 1. Install Core6 with default diskdruid partitioning. 2. Corrupt the LVM 3. Try to boot from the Rescue CD 4. Notice the lack of LVM support. Actual results: No LVM support tools or LVM kernel support. Expected results: LVM support tools and LVM kernel support.
This seems like a gross oversight in the packaging/build of the rescue CD.
I can't reproduce this. I booted an FC6 system with root & home on LVs using the rescue environment's automatic filesystem detection. The LVs were correctly activated and the filesystems mounted at the correct locations: ... /tmp/loop0 on /mnt/runtime type squashfs (ro) /selinux on /selinux type selinuxfs (rw) /dev/t60_vg0/root on /mnt/sysimage type ext3 (rw,data=ordered) /dev/sda1 on /mnt/sysimage/boot type ext3 (rw,data=ordered) /dev/sda5 on /mnt/sysimage/data type ext3 (rw,data=ordered) /dev/t60_vg0/home on /mnt/sysimage/home type ext3 (rw,data=ordered) ... I'm a bit puzzled by the comment that there is no LVM support in the rescue environment - note that only the 'lvm' binary itself is supplied in the rescue environment but this provides all the functionality of the regular lv/vg/pv commands by running 'lvm <command>'. You'll find it at /usr/sbin/lvm (actually /usr is a symlink to /mnt/runtime/usr - a loopback mounted filesystem provided by the stage2.img file). For e.g.: sh-3.1# lvm pvs PV VG Fmt Attr PSize PFree /dev/sda2 t60_vg0 lvm2 a- 14.62G 4.84G sh-3.1# lvm vgdisplay --- Volume group --- VG Name t60_vg0 System ID Format lvm2 Metadata Areas 1 Metadata Sequence No 6 VG Access read/write VG Status resizable MAX LV 0 Cur LV 2 Open LV 2 Max PV 0 Cur PV 1 Act PV 1 VG Size 14.62 GB PE Size 32.00 MB Total PE 468 Alloc PE / Size 313 / 9.78 GB Free PE / Size 155 / 4.84 GB VG UUID Med1jz-abBC-FLPK-xlgX-PdCm-hlR9-hotMRK Once the rescue environment has detected & mounted filesystems, it includes the installed system's bin & sbin directories in $PATH (/mnt/sysimage/sbin etc), so you can just use the normal vgdisplay, pvs, etc. at that point. It would be helpful to know exactly how you 'corrupted' the LVM - I suspect this is the cause of your problems as it looks like it prevented the automatic detection & activation of volume groups in the rescue environment.
Fedora 6 Rescue CD has lvm2 support - I checked i386 and x86_64 and both correctly deteced lvm volumes. So the problem is probably that lvm volumes were seriously currupted. Please if you see another bug (causing lvm corruption) open another bugzilla with more information about it, thanks.