I have a centos 5.6 server that has xen domUs installed on their on
logical volumes. These logical volumes contain their own volume groups
and again their own logical volumes. I want to access the domU logical
volumes and tried this:
[root at kr ~]# fdisk -l /dev/VolGroup00/LogVol02
Disk /dev/VolGroup00/LogVol02: 274.8 GB, 274877906944 bytes
255 heads, 63 sectors/track, 33418 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes
Device Boot Start End Blocks Id System
/dev/VolGroup00/LogVol02p1 * 1 13 104391 83 Linux
/dev/VolGroup00/LogVol02p2 14 16709 134110620
8e Linux LVM
[root at kr ~]# kpartx -a /dev/VolGroup00/LogVol02
[root at kr ~]# ls -l /dev/mapper/
total 0
crw------- 1 root root 10, 62 Apr 26 02:19 control
brw-rw---- 1 root disk 253, 12 Apr 30 00:45 LogVol02p1
brw-rw---- 1 root disk 253, 13 Apr 30 00:45 LogVol02p2
But when I try make the domU volumegroup accessible I run into problems:
[root at kr ~]# vgscan
Reading all physical volumes. This may take a while...
WARNING: Duplicate VG name VolGroup00: Existing
hxNFYw-c1Z6-DB99-SiiK-yRas-eXiI-1w27Z0 (created here) takes precedence
over KmwQTQ-jT19-lvvu-H6d0-i4T3-s5Jw-JHnx0M
WARNING: Duplicate VG name VolGroup00: Existing
hxNFYw-c1Z6-DB99-SiiK-yRas-eXiI-1w27Z0 (created here) takes precedence
over KmwQTQ-jT19-lvvu-H6d0-i4T3-s5Jw-JHnx0M
Found volume group "VolGroup00" using metadata type lvm2
Found volume group "VolGroup00" using metadata type lvm2
How can I now vgchange to the domU volume group as vgchange does not
take the UUID as an argument, just the volume group name...?
Regards,
Peter
On Apr 29, 2011, at 6:05 PM, Peter Peltonen <peter.peltonen at gmail.com> wrote:> I have a centos 5.6 server that has xen domUs installed on their on > logical volumes. These logical volumes contain their own volume groups > and again their own logical volumes. I want to access the domU logical > volumes and tried this: > > [root at kr ~]# fdisk -l /dev/VolGroup00/LogVol02 > > Disk /dev/VolGroup00/LogVol02: 274.8 GB, 274877906944 bytes > 255 heads, 63 sectors/track, 33418 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > > Device Boot Start End Blocks Id System > /dev/VolGroup00/LogVol02p1 * 1 13 104391 83 Linux > /dev/VolGroup00/LogVol02p2 14 16709 134110620 > 8e Linux LVM > > [root at kr ~]# kpartx -a /dev/VolGroup00/LogVol02 > [root at kr ~]# ls -l /dev/mapper/ > total 0 > crw------- 1 root root 10, 62 Apr 26 02:19 control > brw-rw---- 1 root disk 253, 12 Apr 30 00:45 LogVol02p1 > brw-rw---- 1 root disk 253, 13 Apr 30 00:45 LogVol02p2 > > But when I try make the domU volumegroup accessible I run into problems: > > [root at kr ~]# vgscan > Reading all physical volumes. This may take a while... > WARNING: Duplicate VG name VolGroup00: Existing > hxNFYw-c1Z6-DB99-SiiK-yRas-eXiI-1w27Z0 (created here) takes precedence > over KmwQTQ-jT19-lvvu-H6d0-i4T3-s5Jw-JHnx0M > WARNING: Duplicate VG name VolGroup00: Existing > hxNFYw-c1Z6-DB99-SiiK-yRas-eXiI-1w27Z0 (created here) takes precedence > over KmwQTQ-jT19-lvvu-H6d0-i4T3-s5Jw-JHnx0M > Found volume group "VolGroup00" using metadata type lvm2 > Found volume group "VolGroup00" using metadata type lvm2 > > How can I now vgchange to the domU volume group as vgchange does not > take the UUID as an argument, just the volume group name...?That's why I use hostnames instead of generic volume group names. You could vgrename the volume, but the domU won't boot afterwards unless you rename it back and I don't know if it will let you if the vg name already exists. LVM should really use the UUID if the vg name collides or allow it to be imported under a different name temporarily. You could rename your system vg and of course change fstab to match and remake the initrd, then you would be able to load one domU vg at a time. -Ross
On 04/29/11 3:05 PM, Peter Peltonen wrote:> I have a centos 5.6 server that has xen domUs installed on their on > logical volumes. These logical volumes contain their own volume groups > and again their own logical volumes. I want to > ...ugh, and double ugh. this violates the KISS 'keep it super simple' principle, and I can only see the extra layers of complexity leading to more frustration. I would instead run LVM only on dom0, and have your domU's creating file systems directly on the virtual devices mapped to said LVs. if you need to grow a file system, you resize the LV from dom0, and then growfs in the domU