[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: lvm bug - possible solution



On Wed, Aug 22, 2001 at 12:01:14AM +0200, Robert van der Meulen wrote:
> 
> Quoting John H. Robinson, IV (jhriv@ucsd.edu):
> > what about the unexpected reboot? could that cause all the filesystems
> > inside the lvm to become unmountable? (factor filesystem corruption out
> > of the question for this scenario - i am interested only in the lvm part
> > of the equation)
> If you leave 'filesystem corruption' out of the equation, it will give you
> the filesystems as they were the last time 'vgscan' ran.
> 
> If you run 'vgscan' after each change to an lvm volume (or a couple), to
> recreate /etc/lvmtab and /etc/lvmtab.d, it will always be up to date, and
> there will be no problem.

All the LVM commands that change things make a backup of the config files anway,
(unless -An is on the command-line or LVM_AUTOBACKUP=no) so there really
shouldn't be a problem with just running vgchange -ay at boot time. Quite a lot
of people do just that.

If there are troubles then manually booting single-user to run vgscan is the way
around it.
 
patrick
(aka caulfield@sistina.com)



Reply to: