On Sun, 2006-01-01 at 12:47 +0100, Zoltan Boszormenyi wrote:
Florin Andrei írta:
>Another assumption behind XFS is that storage is the best quality and
>therefore it does not lie when it reports back that the data was flushed
>all the way to the magnetic layer (which is something that cheap IDE
>drives/cards lie about, sometimes).
Is IBM/Hitachi Deskstar known to not lie about the flush?
I am almost certain that I've seen a free software tool on the Internet
a while ago that did exactly this: it could tell whether the hard-drive
and/or the controller lie about flushing the data all the way to the
magnetic platter.
I think the technique involved yanking the power cord or hitting the
reset button (different failures, perhaps it's worth testing both) while
the software was doing some complex read/write operations with the disk.
After reboot, the same application looked at the disk and could tell
whether the data was properly flushed or not.
I can't remember the name of the software or anything, but I imagine
that a clever google search would reveal it.
--
Florin Andrei
http://florin.myip.org/