On Thu, Feb 16, 2012 at 6:53 AM, Stefan Sperling <s...@elego.de> wrote:
> >
> > There is no "window of vulnerability".
>
>
> What if the disk controller lies and sends the acknowledgement before
> the actual work has completed (e.g. it cached the write request and is
> going to get to it soon)? How can the OS know what the disk controller
> is really doing?
>
That is a risk in the case of a power failure with a disk/raid controller
that has a cache without battery backup, but in the case of an LVM snapshot
and the power staying on, wouldn't the controller return the cached block
regardless of whether it had made the round trip to disk or not?
--
Les Mikesell
lesmikes...@gmail.com
