Hi,
this behavior has a unexpected side effect if you use pvmove and the
system gets rebooted while the mirroring is still in progress:
- run pvmove to move one pv to another (does not mather if the pv is
part of the root vg or not)
- reboot the system
- the initrd will hang due it could not find the root volume (but the
mirror process is started by the lvm initrd scripts in the background)
- the system could recoverd by:
- specify a root=/dev/mapper/... parameter on the lilo command line
- wait 'til the pvmove process has completed and reboot
If pvmove is used to have a zero downtime migration of one pv to another
pv it might become a real downtime if the system is rebooted (i.e.
powerloss) during the process.
Regards,
Thomas Liske
--
To UNSUBSCRIBE, email to [EMAIL PROTECTED]
with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
