On Mon, 2003-05-12 at 05:23, Axel Thimm wrote:
> I know about XFS's zeroing of files for such files that were still marked as
> dirty when the crash occured. I wonder about two things:
> o Why are 15-20 minutes old files also wiped out? Are the write-backs that
> much delayed (even if there is not other FS/CPU load)?
> o What is the best way to find these files (why aren't their names simply
> dumped to the kernel logs)? Something like
> find . -xdev -type f \! -empty | xargs <grep for non-zero character here>
> What's the best expression for the last part?
> (Red Hat 8.0 & XFS 1.2)
You need to try the current cvs kernel, there was a major rework of xfs
sync recently which should have fixed this.