Procedure to deactivate vg00 during an active login without the "in use" error.
In HP-UX you can boot the system into LVM Maintenance mode
in order to perform lvm maintenance on vg00. I am seeking an equivalent process via Redhat Linux Enterprise Servers, so I would appreciate it very much if someone can post the exact procedure to accomplish this function. Thank You!!! The exact maintenance that I am trying to perform is to restore the PVID on the PV associated with VG00. /sbin/pvscan PV /dev/hda2 VG vg00 lvm2 [37.19 GB / 22.62 GB free] Total: 1 [37.19 GB] / in use: 1 [37.19 GB] / in no VG: 0 [0 ] When I try to execute the "pvcreate" command to restore the PVID on /dev/hda2 I get the following results. Really INITIALIZE physical volume "/dev/hda2" of volume group "vg00" [y/n]? y Can't open /dev/hda2 exclusively. Mounted filesystem? This is a "catch-22" since I would have to "umount" all filesystems associated with vg00 which includes "/usr" where the actual "pvcreate" command resides and "/etc under /" where the "vg00_00000.vg" config file exists. So the question is still basically the same as I am looking for the actual proceedure to accomplish this task? |
Need more info. :) What is the exact maintenance you're looking to perform on vg00 (presumably a volume group)? Most maintenance you can perform on a live system; there are exceptions, like resizing the logical volume that is mounted to /.
|
# umount -l /dev/vg00/lvm2
Proceed with your PV commands. Then proceed with what you're doing - please note, if you're trying to bring down / up a VG that you're actively using - that's not recommended. In most cases dealing with a live environment can be a pain - you may need to schedule a downtime in order to boot your system into runlevel 1; or temporarily use a rescue environment. |
All times are GMT -5. The time now is 12:38 AM. |